id	sid	tid	token	lemma	pos
esrj-89750	1	1	keywords	keyword	NOUN
esrj-89750	1	2	:	:	PUNCT
esrj-89750	1	3	deep	deep	ADJ
esrj-89750	1	4	learning	learning	NOUN
esrj-89750	1	5	;	;	PUNCT
esrj-89750	1	6	soil	soil	NOUN
esrj-89750	1	7	remote	remote	ADJ
esrj-89750	1	8	sensing	sense	VERB
esrj-89750	1	9	image	image	NOUN
esrj-89750	1	10	;	;	PUNCT
esrj-89750	1	11	maximum	maximum	ADJ
esrj-89750	1	12	likelihood	likelihood	NOUN
esrj-89750	1	13	estimation	estimation	NOUN
esrj-89750	1	14	;	;	PUNCT
esrj-89750	1	15	classification	classification	NOUN
esrj-89750	1	16	method	method	NOUN
esrj-89750	1	17	.	.	PUNCT
esrj-89750	2	1	palabras	palabras	PROPN
esrj-89750	2	2	clave	clave	PROPN
esrj-89750	2	3	:	:	PUNCT
esrj-89750	2	4	aprendizaje	aprendizaje	PROPN
esrj-89750	2	5	profundo	profundo	PROPN
esrj-89750	2	6	;	;	PUNCT
esrj-89750	2	7	imagen	imagen	PROPN
esrj-89750	2	8	de	de	PROPN
esrj-89750	2	9	teledetección	teledetección	PROPN
esrj-89750	2	10	del	del	PROPN
esrj-89750	2	11	suelo	suelo	NOUN
esrj-89750	2	12	;	;	PUNCT
esrj-89750	2	13	estimación	estimación	X
esrj-89750	2	14	de	de	X
esrj-89750	2	15	máxima	máxima	X
esrj-89750	2	16	verosimilitud	verosimilitud	NOUN
esrj-89750	2	17	;	;	PUNCT
esrj-89750	2	18	método	método	X
esrj-89750	2	19	de	de	X
esrj-89750	2	20	clasificación	clasificación	PROPN
esrj-89750	2	21	.	.	PUNCT
esrj-89750	3	1	how	how	SCONJ
esrj-89750	3	2	to	to	PART
esrj-89750	3	3	cite	cite	VERB
esrj-89750	3	4	item	item	PROPN
esrj-89750	3	5	liang	liang	PROPN
esrj-89750	3	6	,	,	PUNCT
esrj-89750	3	7	s.	s.	PROPN
esrj-89750	3	8	,	,	PUNCT
esrj-89750	3	9	cheng	cheng	PROPN
esrj-89750	3	10	,	,	PUNCT
esrj-89750	3	11	j.	j.	PROPN
esrj-89750	3	12	,	,	PUNCT
esrj-89750	3	13	&	&	CCONJ
esrj-89750	3	14	zhang	zhang	PROPN
esrj-89750	3	15	,	,	PUNCT
esrj-89750	3	16	j.	j.	PROPN
esrj-89750	3	17	(	(	PUNCT
esrj-89750	3	18	2020	2020	NUM
esrj-89750	3	19	)	)	PUNCT
esrj-89750	3	20	.	.	PUNCT
esrj-89750	4	1	maximum	maximum	ADJ
esrj-89750	4	2	likelihood	likelihood	NOUN
esrj-89750	4	3	classification	classification	NOUN
esrj-89750	4	4	of	of	ADP
esrj-89750	4	5	soil	soil	NOUN
esrj-89750	4	6	remote	remote	ADJ
esrj-89750	4	7	sensing	sense	VERB
esrj-89750	4	8	image	image	NOUN
esrj-89750	4	9	based	base	VERB
esrj-89750	4	10	on	on	ADP
esrj-89750	4	11	deep	deep	ADJ
esrj-89750	4	12	learning	learning	NOUN
esrj-89750	4	13	.	.	PUNCT
esrj-89750	5	1	earth	earth	PROPN
esrj-89750	5	2	sciences	sciences	PROPN
esrj-89750	5	3	research	research	PROPN
esrj-89750	5	4	journal	journal	PROPN
esrj-89750	5	5	,	,	PUNCT
esrj-89750	5	6	24(3	24(3	NUM
esrj-89750	5	7	)	)	PUNCT
esrj-89750	5	8	,	,	PUNCT
esrj-89750	5	9	357	357	NUM
esrj-89750	5	10	-	-	SYM
esrj-89750	5	11	365	365	NUM
esrj-89750	5	12	.	.	PUNCT
esrj-89750	6	1	doi	doi	NOUN
esrj-89750	6	2	:	:	PUNCT
esrj-89750	6	3	https://doi.org/10.15446/esrj	https://doi.org/10.15446/esrj	NOUN
esrj-89750	6	4	.	.	PUNCT
esrj-89750	7	1	v24n3.89750	v24n3.89750	ADP
esrj-89750	7	2	soil	soil	NOUN
esrj-89750	7	3	remote	remote	ADJ
esrj-89750	7	4	sensing	sense	VERB
esrj-89750	7	5	image	image	NOUN
esrj-89750	7	6	classification	classification	NOUN
esrj-89750	7	7	is	be	AUX
esrj-89750	7	8	the	the	DET
esrj-89750	7	9	most	most	ADV
esrj-89750	7	10	difficult	difficult	ADJ
esrj-89750	7	11	in	in	ADP
esrj-89750	7	12	the	the	DET
esrj-89750	7	13	national	national	ADJ
esrj-89750	7	14	soil	soil	NOUN
esrj-89750	7	15	census	census	NOUN
esrj-89750	7	16	work	work	NOUN
esrj-89750	7	17	.	.	PUNCT
esrj-89750	8	1	current	current	ADJ
esrj-89750	8	2	soil	soil	NOUN
esrj-89750	8	3	remote	remote	ADJ
esrj-89750	8	4	sensing	sense	VERB
esrj-89750	8	5	image	image	NOUN
esrj-89750	8	6	classification	classification	NOUN
esrj-89750	8	7	methods	method	NOUN
esrj-89750	8	8	based	base	VERB
esrj-89750	8	9	on	on	ADP
esrj-89750	8	10	deep	deep	ADJ
esrj-89750	8	11	learning	learning	NOUN
esrj-89750	8	12	and	and	CCONJ
esrj-89750	8	13	maximum	maximum	ADJ
esrj-89750	8	14	likelihood	likelihood	NOUN
esrj-89750	8	15	estimation	estimation	NOUN
esrj-89750	8	16	are	be	AUX
esrj-89750	8	17	challenging	challenge	VERB
esrj-89750	8	18	to	to	PART
esrj-89750	8	19	meet	meet	VERB
esrj-89750	8	20	the	the	DET
esrj-89750	8	21	actual	actual	ADJ
esrj-89750	8	22	needs	need	NOUN
esrj-89750	8	23	.	.	PUNCT
esrj-89750	9	1	therefore	therefore	ADV
esrj-89750	9	2	,	,	PUNCT
esrj-89750	9	3	this	this	DET
esrj-89750	9	4	paper	paper	NOUN
esrj-89750	9	5	combines	combine	VERB
esrj-89750	9	6	deep	deep	ADJ
esrj-89750	9	7	learning	learning	NOUN
esrj-89750	9	8	with	with	ADP
esrj-89750	9	9	maximum	maximum	ADJ
esrj-89750	9	10	likelihood	likelihood	NOUN
esrj-89750	9	11	estimation	estimation	NOUN
esrj-89750	9	12	and	and	CCONJ
esrj-89750	9	13	proposes	propose	VERB
esrj-89750	9	14	a	a	DET
esrj-89750	9	15	maximum	maximum	ADJ
esrj-89750	9	16	likelihood	likelihood	NOUN
esrj-89750	9	17	classification	classification	NOUN
esrj-89750	9	18	method	method	NOUN
esrj-89750	9	19	for	for	ADP
esrj-89750	9	20	soil	soil	NOUN
esrj-89750	9	21	remote	remote	ADJ
esrj-89750	9	22	sensing	sensing	NOUN
esrj-89750	9	23	images	image	NOUN
esrj-89750	9	24	based	base	VERB
esrj-89750	9	25	on	on	ADP
esrj-89750	9	26	deep	deep	ADJ
esrj-89750	9	27	learning	learning	NOUN
esrj-89750	9	28	.	.	PUNCT
esrj-89750	10	1	the	the	DET
esrj-89750	10	2	method	method	NOUN
esrj-89750	10	3	is	be	AUX
esrj-89750	10	4	divided	divide	VERB
esrj-89750	10	5	into	into	ADP
esrj-89750	10	6	four	four	NUM
esrj-89750	10	7	parts	part	NOUN
esrj-89750	10	8	.	.	PUNCT
esrj-89750	11	1	firstly	firstly	ADV
esrj-89750	11	2	,	,	PUNCT
esrj-89750	11	3	the	the	DET
esrj-89750	11	4	pretreatment	pretreatment	NOUN
esrj-89750	11	5	of	of	ADP
esrj-89750	11	6	soil	soil	NOUN
esrj-89750	11	7	remote	remote	ADJ
esrj-89750	11	8	sensing	sense	VERB
esrj-89750	11	9	image	image	NOUN
esrj-89750	11	10	is	be	AUX
esrj-89750	11	11	carried	carry	VERB
esrj-89750	11	12	out	out	ADP
esrj-89750	11	13	,	,	PUNCT
esrj-89750	11	14	including	include	VERB
esrj-89750	11	15	three	three	NUM
esrj-89750	11	16	processes	process	NOUN
esrj-89750	11	17	:	:	PUNCT
esrj-89750	11	18	image	image	NOUN
esrj-89750	11	19	gray	gray	ADJ
esrj-89750	11	20	,	,	PUNCT
esrj-89750	11	21	image	image	NOUN
esrj-89750	11	22	denoising	denoising	NOUN
esrj-89750	11	23	,	,	PUNCT
esrj-89750	11	24	and	and	CCONJ
esrj-89750	11	25	image	image	NOUN
esrj-89750	11	26	correction	correction	NOUN
esrj-89750	11	27	;	;	PUNCT
esrj-89750	11	28	secondly	secondly	ADV
esrj-89750	11	29	,	,	PUNCT
esrj-89750	11	30	the	the	DET
esrj-89750	11	31	target	target	NOUN
esrj-89750	11	32	of	of	ADP
esrj-89750	11	33	soil	soil	NOUN
esrj-89750	11	34	remote	remote	ADJ
esrj-89750	11	35	sensing	sense	VERB
esrj-89750	11	36	image	image	NOUN
esrj-89750	11	37	is	be	AUX
esrj-89750	11	38	detected	detect	VERB
esrj-89750	11	39	by	by	ADP
esrj-89750	11	40	deep	deep	ADJ
esrj-89750	11	41	learning	learning	NOUN
esrj-89750	11	42	algorithm	algorithm	NOUN
esrj-89750	11	43	;	;	PUNCT
esrj-89750	11	44	thirdly	thirdly	ADV
esrj-89750	11	45	,	,	PUNCT
esrj-89750	11	46	the	the	DET
esrj-89750	11	47	maximum	maximum	ADJ
esrj-89750	11	48	likelihood	likelihood	NOUN
esrj-89750	11	49	algorithm	algorithm	NOUN
esrj-89750	11	50	is	be	AUX
esrj-89750	11	51	used	use	VERB
esrj-89750	11	52	to	to	PART
esrj-89750	11	53	classify	classify	VERB
esrj-89750	11	54	soil	soil	NOUN
esrj-89750	11	55	remote	remote	ADJ
esrj-89750	11	56	sensing	sense	VERB
esrj-89750	11	57	image	image	NOUN
esrj-89750	11	58	;	;	PUNCT
esrj-89750	11	59	finally	finally	ADV
esrj-89750	11	60	,	,	PUNCT
esrj-89750	11	61	the	the	DET
esrj-89750	11	62	classification	classification	NOUN
esrj-89750	11	63	performance	performance	NOUN
esrj-89750	11	64	is	be	AUX
esrj-89750	11	65	tested	test	VERB
esrj-89750	11	66	by	by	ADP
esrj-89750	11	67	an	an	DET
esrj-89750	11	68	example	example	NOUN
esrj-89750	11	69	.	.	PUNCT
esrj-89750	12	1	the	the	DET
esrj-89750	12	2	results	result	NOUN
esrj-89750	12	3	show	show	VERB
esrj-89750	12	4	that	that	SCONJ
esrj-89750	12	5	this	this	DET
esrj-89750	12	6	method	method	NOUN
esrj-89750	12	7	can	can	AUX
esrj-89750	12	8	effectively	effectively	ADV
esrj-89750	12	9	segment	segment	VERB
esrj-89750	12	10	the	the	DET
esrj-89750	12	11	remote	remote	ADJ
esrj-89750	12	12	sensing	sense	VERB
esrj-89750	12	13	image	image	NOUN
esrj-89750	12	14	of	of	ADP
esrj-89750	12	15	soil	soil	NOUN
esrj-89750	12	16	,	,	PUNCT
esrj-89750	12	17	and	and	CCONJ
esrj-89750	12	18	the	the	DET
esrj-89750	12	19	segmentation	segmentation	NOUN
esrj-89750	12	20	accuracy	accuracy	NOUN
esrj-89750	12	21	is	be	AUX
esrj-89750	12	22	high	high	ADJ
esrj-89750	12	23	,	,	PUNCT
esrj-89750	12	24	which	which	PRON
esrj-89750	12	25	proves	prove	VERB
esrj-89750	12	26	the	the	DET
esrj-89750	12	27	effectiveness	effectiveness	NOUN
esrj-89750	12	28	and	and	CCONJ
esrj-89750	12	29	superiority	superiority	NOUN
esrj-89750	12	30	of	of	ADP
esrj-89750	12	31	the	the	DET
esrj-89750	12	32	method	method	NOUN
esrj-89750	12	33	.	.	PUNCT
esrj-89750	13	1	abstract	abstract	ADJ
esrj-89750	13	2	maximum	maximum	ADJ
esrj-89750	13	3	likelihood	likelihood	NOUN
esrj-89750	13	4	classification	classification	NOUN
esrj-89750	13	5	of	of	ADP
esrj-89750	13	6	soil	soil	NOUN
esrj-89750	13	7	remote	remote	ADJ
esrj-89750	13	8	sensing	sense	VERB
esrj-89750	13	9	image	image	NOUN
esrj-89750	13	10	based	base	VERB
esrj-89750	13	11	on	on	ADP
esrj-89750	13	12	deep	deep	ADJ
esrj-89750	13	13	learning	learn	VERB
esrj-89750	13	14	clasificación	clasificación	PROPN
esrj-89750	13	15	de	de	PROPN
esrj-89750	13	16	verosimilitud	verosimilitud	PROPN
esrj-89750	13	17	máxima	máxima	PROPN
esrj-89750	13	18	de	de	PROPN
esrj-89750	13	19	imágenes	imágenes	PROPN
esrj-89750	13	20	en	en	ADP
esrj-89750	13	21	teledetección	teledetección	PROPN
esrj-89750	13	22	del	del	PROPN
esrj-89750	13	23	suelo	suelo	PROPN
esrj-89750	13	24	con	con	PROPN
esrj-89750	13	25	base	base	PROPN
esrj-89750	13	26	en	en	PROPN
esrj-89750	13	27	aprendizaje	aprendizaje	PROPN
esrj-89750	13	28	profundo	profundo	PROPN
esrj-89750	13	29	automático	automático	PROPN
esrj-89750	13	30	issn	issn	PROPN
esrj-89750	13	31	1794	1794	NUM
esrj-89750	13	32	-	-	SYM
esrj-89750	13	33	6190	6190	NUM
esrj-89750	13	34	e	e	NOUN
esrj-89750	13	35	-	-	NOUN
esrj-89750	13	36	issn	issn	PROPN
esrj-89750	13	37	2339	2339	NUM
esrj-89750	13	38	-	-	SYM
esrj-89750	13	39	3459	3459	NUM
esrj-89750	13	40	https://doi.org/10.15446/esrj.v24n3.89750	https://doi.org/10.15446/esrj.v24n3.89750	NOUN
esrj-89750	13	41	la	la	PROPN
esrj-89750	13	42	clasificación	clasificación	PROPN
esrj-89750	13	43	de	de	PROPN
esrj-89750	13	44	imágenes	imágenes	PROPN
esrj-89750	13	45	de	de	PROPN
esrj-89750	13	46	detección	detección	PROPN
esrj-89750	13	47	remota	remota	PROPN
esrj-89750	13	48	de	de	X
esrj-89750	13	49	suelos	suelos	PROPN
esrj-89750	13	50	es	es	PROPN
esrj-89750	13	51	la	la	PROPN
esrj-89750	13	52	más	más	PROPN
esrj-89750	13	53	difícil	difícil	PROPN
esrj-89750	13	54	en	en	PROPN
esrj-89750	13	55	el	el	PROPN
esrj-89750	13	56	trabajo	trabajo	PROPN
esrj-89750	13	57	del	del	PROPN
esrj-89750	13	58	censo	censo	PROPN
esrj-89750	13	59	nacional	nacional	PROPN
esrj-89750	13	60	de	de	PROPN
esrj-89750	13	61	suelos	suelos	PROPN
esrj-89750	13	62	en	en	PROPN
esrj-89750	13	63	china	china	PROPN
esrj-89750	13	64	.	.	PUNCT
esrj-89750	14	1	los	los	PROPN
esrj-89750	14	2	métodos	métodos	PROPN
esrj-89750	14	3	vigentes	vigentes	PROPN
esrj-89750	14	4	de	de	PROPN
esrj-89750	14	5	clasificación	clasificación	PROPN
esrj-89750	14	6	de	de	PROPN
esrj-89750	14	7	imágenes	imágenes	PROPN
esrj-89750	14	8	de	de	PROPN
esrj-89750	14	9	teledetección	teledetección	PROPN
esrj-89750	14	10	del	del	PROPN
esrj-89750	14	11	suelo	suelo	NOUN
esrj-89750	14	12	basados	basado	VERB
esrj-89750	14	13	en	en	PROPN
esrj-89750	14	14	el	el	PROPN
esrj-89750	14	15	aprendizaje	aprendizaje	PROPN
esrj-89750	14	16	profundo	profundo	PROPN
esrj-89750	14	17	y	y	PROPN
esrj-89750	14	18	la	la	PROPN
esrj-89750	14	19	estimación	estimación	PROPN
esrj-89750	14	20	de	de	ADP
esrj-89750	14	21	máxima	máxima	PROPN
esrj-89750	14	22	probabilidad	probabilidad	X
esrj-89750	14	23	no	no	DET
esrj-89750	14	24	satisfacen	satisfacen	PROPN
esrj-89750	14	25	las	las	PROPN
esrj-89750	14	26	necesidades	necesidades	PROPN
esrj-89750	14	27	actuales	actuales	PROPN
esrj-89750	14	28	.	.	PUNCT
esrj-89750	15	1	por	por	NOUN
esrj-89750	15	2	lo	lo	PROPN
esrj-89750	15	3	tanto	tanto	NOUN
esrj-89750	15	4	,	,	PUNCT
esrj-89750	15	5	este	este	X
esrj-89750	15	6	documento	documento	PROPN
esrj-89750	15	7	combina	combina	PROPN
esrj-89750	15	8	el	el	PROPN
esrj-89750	15	9	aprendizaje	aprendizaje	PROPN
esrj-89750	15	10	profundo	profundo	PROPN
esrj-89750	15	11	con	con	PROPN
esrj-89750	15	12	la	la	X
esrj-89750	15	13	estimación	estimación	PROPN
esrj-89750	15	14	de	de	X
esrj-89750	15	15	máxima	máxima	PROPN
esrj-89750	15	16	verosimilitud	verosimilitud	PROPN
esrj-89750	15	17	y	y	PROPN
esrj-89750	15	18	propone	propone	PROPN
esrj-89750	15	19	un	un	PROPN
esrj-89750	15	20	método	método	PROPN
esrj-89750	15	21	de	de	PROPN
esrj-89750	15	22	clasificación	clasificación	PROPN
esrj-89750	15	23	para	para	PROPN
esrj-89750	15	24	estas	estas	PROPN
esrj-89750	15	25	imágenes	imágenes	PROPN
esrj-89750	15	26	de	de	PROPN
esrj-89750	15	27	teledetección	teledetección	PROPN
esrj-89750	15	28	.	.	PUNCT
esrj-89750	16	1	en	en	PROPN
esrj-89750	16	2	primer	primer	PROPN
esrj-89750	16	3	lugar	lugar	PROPN
esrj-89750	16	4	,	,	PUNCT
esrj-89750	16	5	se	se	PROPN
esrj-89750	16	6	lleva	lleva	PROPN
esrj-89750	16	7	a	a	DET
esrj-89750	16	8	cabo	cabo	NOUN
esrj-89750	16	9	el	el	PROPN
esrj-89750	16	10	preprocesamiento	preprocesamiento	PROPN
esrj-89750	16	11	de	de	X
esrj-89750	16	12	la	la	X
esrj-89750	16	13	imagen	imagen	PROPN
esrj-89750	16	14	de	de	PROPN
esrj-89750	16	15	teledetección	teledetección	PROPN
esrj-89750	16	16	del	del	X
esrj-89750	16	17	suelo	suelo	NOUN
esrj-89750	16	18	,	,	PUNCT
esrj-89750	16	19	lo	lo	PROPN
esrj-89750	16	20	que	que	PROPN
esrj-89750	16	21	incluye	incluye	PROPN
esrj-89750	16	22	tres	tres	X
esrj-89750	16	23	procesos	procesos	PROPN
esrj-89750	16	24	:	:	PUNCT
esrj-89750	16	25	imagen	imagen	PROPN
esrj-89750	16	26	gris	gris	PROPN
esrj-89750	16	27	,	,	PUNCT
esrj-89750	16	28	eliminación	eliminación	ADP
esrj-89750	16	29	de	de	PROPN
esrj-89750	16	30	ruido	ruido	PROPN
esrj-89750	16	31	y	y	PROPN
esrj-89750	16	32	corrección	corrección	PROPN
esrj-89750	16	33	de	de	X
esrj-89750	16	34	imagen	imagen	PROPN
esrj-89750	16	35	;	;	PUNCT
esrj-89750	16	36	en	en	X
esrj-89750	16	37	segundo	segundo	PROPN
esrj-89750	16	38	lugar	lugar	PROPN
esrj-89750	16	39	,	,	PUNCT
esrj-89750	16	40	el	el	PROPN
esrj-89750	16	41	objetivo	objetivo	PROPN
esrj-89750	16	42	de	de	PROPN
esrj-89750	16	43	la	la	PROPN
esrj-89750	16	44	imagen	imagen	PROPN
esrj-89750	16	45	del	del	PROPN
esrj-89750	16	46	suelo	suelo	PROPN
esrj-89750	16	47	se	se	PROPN
esrj-89750	16	48	detecta	detecta	PROPN
esrj-89750	16	49	mediante	mediante	PROPN
esrj-89750	16	50	un	un	PROPN
esrj-89750	16	51	algoritmo	algoritmo	PROPN
esrj-89750	16	52	de	de	PROPN
esrj-89750	16	53	aprendizaje	aprendizaje	PROPN
esrj-89750	16	54	profundo	profundo	PROPN
esrj-89750	16	55	;	;	PUNCT
esrj-89750	16	56	tercero	tercero	PROPN
esrj-89750	16	57	,	,	PUNCT
esrj-89750	16	58	el	el	PROPN
esrj-89750	16	59	algoritmo	algoritmo	PROPN
esrj-89750	16	60	de	de	X
esrj-89750	16	61	máxima	máxima	PROPN
esrj-89750	16	62	verosimilitud	verosimilitud	PROPN
esrj-89750	16	63	se	se	PROPN
esrj-89750	16	64	usa	usa	PROPN
esrj-89750	16	65	para	para	PROPN
esrj-89750	16	66	clasificar	clasificar	PROPN
esrj-89750	16	67	la	la	PROPN
esrj-89750	16	68	imagen	imagen	PROPN
esrj-89750	16	69	de	de	PROPN
esrj-89750	16	70	detección	detección	PROPN
esrj-89750	16	71	remota	remota	X
esrj-89750	16	72	del	del	X
esrj-89750	16	73	suelo	suelo	PROPN
esrj-89750	16	74	;	;	PUNCT
esrj-89750	16	75	y	y	PROPN
esrj-89750	16	76	,	,	PUNCT
esrj-89750	16	77	finalmente	finalmente	PROPN
esrj-89750	16	78	,	,	PUNCT
esrj-89750	16	79	el	el	PROPN
esrj-89750	16	80	rendimiento	rendimiento	PROPN
esrj-89750	16	81	de	de	PROPN
esrj-89750	16	82	la	la	PROPN
esrj-89750	16	83	clasificación	clasificación	PROPN
esrj-89750	16	84	se	se	PROPN
esrj-89750	16	85	prueba	prueba	PROPN
esrj-89750	16	86	con	con	PROPN
esrj-89750	16	87	un	un	PROPN
esrj-89750	16	88	ejemplo	ejemplo	PROPN
esrj-89750	16	89	.	.	PUNCT
esrj-89750	17	1	los	los	PROPN
esrj-89750	17	2	resultados	resultado	VERB
esrj-89750	17	3	muestran	muestran	ADJ
esrj-89750	17	4	que	que	X
esrj-89750	17	5	este	este	X
esrj-89750	17	6	método	método	PROPN
esrj-89750	17	7	puede	puede	PROPN
esrj-89750	17	8	segmentar	segmentar	PROPN
esrj-89750	17	9	efectivamente	efectivamente	PROPN
esrj-89750	17	10	la	la	PROPN
esrj-89750	17	11	imagen	imagen	PROPN
esrj-89750	17	12	de	de	PROPN
esrj-89750	17	13	detección	detección	PROPN
esrj-89750	17	14	remota	remota	X
esrj-89750	17	15	del	del	PROPN
esrj-89750	17	16	suelo	suelo	PROPN
esrj-89750	17	17	,	,	PUNCT
esrj-89750	17	18	y	y	PROPN
esrj-89750	17	19	la	la	PROPN
esrj-89750	17	20	precisión	precisión	X
esrj-89750	17	21	de	de	PROPN
esrj-89750	17	22	la	la	PROPN
esrj-89750	17	23	segmentación	segmentación	PROPN
esrj-89750	17	24	es	es	X
esrj-89750	17	25	alta	alta	PROPN
esrj-89750	17	26	,	,	PUNCT
esrj-89750	17	27	lo	lo	PROPN
esrj-89750	17	28	que	que	PROPN
esrj-89750	17	29	demuestra	demuestra	PROPN
esrj-89750	17	30	la	la	PROPN
esrj-89750	17	31	efectividad	efectividad	PROPN
esrj-89750	17	32	y	y	PROPN
esrj-89750	17	33	superioridad	superioridad	PROPN
esrj-89750	17	34	del	del	PROPN
esrj-89750	17	35	método	método	PROPN
esrj-89750	17	36	.	.	PUNCT
esrj-89750	18	1	resumen	resumen	PROPN
esrj-89750	18	2	record	record	PROPN
esrj-89750	18	3	manuscript	manuscript	NOUN
esrj-89750	18	4	received	receive	VERB
esrj-89750	18	5	:	:	PUNCT
esrj-89750	18	6	20/05/2019	20/05/2019	NUM
esrj-89750	18	7	accepted	accept	VERB
esrj-89750	18	8	for	for	ADP
esrj-89750	18	9	publication	publication	NOUN
esrj-89750	18	10	:	:	PUNCT
esrj-89750	18	11	11/03/2020	11/03/2020	NUM
esrj-89750	18	12	earth	earth	NOUN
esrj-89750	18	13	sciences	sciences	PROPN
esrj-89750	18	14	research	research	PROPN
esrj-89750	18	15	journal	journal	PROPN
esrj-89750	18	16	earth	earth	PROPN
esrj-89750	18	17	sci	sci	PROPN
esrj-89750	18	18	.	.	PUNCT
esrj-89750	19	1	res	res	PROPN
esrj-89750	19	2	.	.	PUNCT
esrj-89750	20	1	j.	j.	PROPN
esrj-89750	20	2	vol	vol	PROPN
esrj-89750	20	3	.	.	PROPN
esrj-89750	21	1	24	24	NUM
esrj-89750	21	2	,	,	PUNCT
esrj-89750	21	3	no	no	INTJ
esrj-89750	21	4	.	.	NOUN
esrj-89750	21	5	3	3	NUM
esrj-89750	21	6	(	(	PUNCT
esrj-89750	21	7	september	september	PROPN
esrj-89750	21	8	,	,	PUNCT
esrj-89750	21	9	2020	2020	NUM
esrj-89750	21	10	):	):	PUNCT
esrj-89750	21	11	357	357	NUM
esrj-89750	21	12	-	-	SYM
esrj-89750	21	13	365	365	NUM
esrj-89750	21	14	shujun	shujun	PROPN
esrj-89750	21	15	liang1	liang1	PROPN
esrj-89750	21	16	,	,	PUNCT
esrj-89750	21	17	jing	je	VERB
esrj-89750	22	1	cheng2	cheng2	PROPN
esrj-89750	22	2	*	*	PROPN
esrj-89750	22	3	,	,	PUNCT
esrj-89750	22	4	jianwei	jianwei	PROPN
esrj-89750	22	5	zhang1	zhang1	PROPN
esrj-89750	23	1	1software	1software	NUM
esrj-89750	23	2	engineering	engineering	NOUN
esrj-89750	23	3	college	college	NOUN
esrj-89750	23	4	,	,	PUNCT
esrj-89750	23	5	zhengzhou	zhengzhou	PROPN
esrj-89750	23	6	university	university	PROPN
esrj-89750	23	7	of	of	ADP
esrj-89750	23	8	light	light	ADJ
esrj-89750	23	9	industry	industry	NOUN
esrj-89750	23	10	,	,	PUNCT
esrj-89750	23	11	zhengzhou	zhengzhou	PROPN
esrj-89750	23	12	450002	450002	NUM
esrj-89750	23	13	,	,	PUNCT
esrj-89750	23	14	china	china	PROPN
esrj-89750	23	15	2engineering	2engineering	NUM
esrj-89750	23	16	training	training	NOUN
esrj-89750	23	17	center	center	NOUN
esrj-89750	23	18	,	,	PUNCT
esrj-89750	23	19	zhengzhou	zhengzhou	PROPN
esrj-89750	23	20	university	university	PROPN
esrj-89750	23	21	of	of	ADP
esrj-89750	23	22	light	light	ADJ
esrj-89750	23	23	industry	industry	NOUN
esrj-89750	23	24	,	,	PUNCT
esrj-89750	23	25	zhengzhou	zhengzhou	PROPN
esrj-89750	23	26	450002	450002	NUM
esrj-89750	23	27	,	,	PUNCT
esrj-89750	23	28	china	china	PROPN
esrj-89750	23	29	*	*	PUNCT
esrj-89750	23	30	corresponding	correspond	VERB
esrj-89750	23	31	author	author	NOUN
esrj-89750	23	32	:	:	PUNCT
esrj-89750	23	33	zzqlcy126@163.com	zzqlcy126@163.com	X
esrj-89750	23	34	re	re	VERB
esrj-89750	23	35	m	m	PROPN
esrj-89750	23	36	o	o	VERB
esrj-89750	23	37	te	te	X
esrj-89750	23	38	s	s	X
esrj-89750	23	39	en	en	ADP
esrj-89750	23	40	si	si	PROPN
esrj-89750	23	41	ng	ng	PROPN
esrj-89750	23	42	https://doi.org/10.15446/esrj.v24n3.89750	https://doi.org/10.15446/esrj.v24n3.89750	X
esrj-89750	23	43	https://doi.org/10.15446/esrj.v24n3.89750	https://doi.org/10.15446/esrj.v24n3.89750	X
esrj-89750	23	44	https://doi.org/10.15446/esrj.v24n3.89750	https://doi.org/10.15446/esrj.v24n3.89750	X
esrj-89750	23	45	mailto:zzqlcy126@163.com	mailto:zzqlcy126@163.com	PROPN
esrj-89750	23	46	358	358	NUM
esrj-89750	23	47	shujun	shujun	PROPN
esrj-89750	23	48	liang1	liang1	PROPN
esrj-89750	23	49	,	,	PUNCT
esrj-89750	23	50	jing	je	VERB
esrj-89750	23	51	cheng2	cheng2	PROPN
esrj-89750	23	52	*	*	PROPN
esrj-89750	23	53	,	,	PUNCT
esrj-89750	23	54	jianwei	jianwei	PROPN
esrj-89750	23	55	zhang1	zhang1	PROPN
esrj-89750	23	56	introduction	introduction	NOUN
esrj-89750	23	57	china	china	PROPN
esrj-89750	23	58	has	have	VERB
esrj-89750	23	59	a	a	DET
esrj-89750	23	60	vast	vast	ADJ
esrj-89750	23	61	territory	territory	NOUN
esrj-89750	23	62	and	and	CCONJ
esrj-89750	23	63	abundant	abundant	ADJ
esrj-89750	23	64	land	land	NOUN
esrj-89750	23	65	resources	resource	NOUN
esrj-89750	23	66	.	.	PUNCT
esrj-89750	24	1	its	its	PRON
esrj-89750	24	2	total	total	ADJ
esrj-89750	24	3	area	area	NOUN
esrj-89750	24	4	is	be	AUX
esrj-89750	24	5	about	about	ADV
esrj-89750	24	6	9.6	9.6	NUM
esrj-89750	24	7	million	million	NUM
esrj-89750	24	8	square	square	ADJ
esrj-89750	24	9	kilometers	kilometer	NOUN
esrj-89750	24	10	,	,	PUNCT
esrj-89750	24	11	second	second	ADJ
esrj-89750	24	12	only	only	ADV
esrj-89750	24	13	to	to	ADP
esrj-89750	24	14	russia	russia	PROPN
esrj-89750	24	15	and	and	CCONJ
esrj-89750	24	16	canada	canada	PROPN
esrj-89750	24	17	,	,	PUNCT
esrj-89750	24	18	ranking	rank	VERB
esrj-89750	24	19	third	third	ADV
esrj-89750	24	20	globally	globally	ADV
esrj-89750	24	21	.	.	PUNCT
esrj-89750	25	1	two	two	NUM
esrj-89750	25	2	national	national	ADJ
esrj-89750	25	3	soil	soil	NOUN
esrj-89750	25	4	surveys	survey	NOUN
esrj-89750	25	5	have	have	AUX
esrj-89750	25	6	been	be	AUX
esrj-89750	25	7	carried	carry	VERB
esrj-89750	25	8	out	out	ADP
esrj-89750	25	9	from	from	ADP
esrj-89750	25	10	1958	1958	NUM
esrj-89750	25	11	to	to	ADP
esrj-89750	25	12	1960	1960	NUM
esrj-89750	25	13	and	and	CCONJ
esrj-89750	25	14	1979	1979	NUM
esrj-89750	25	15	to	to	ADP
esrj-89750	25	16	1985	1985	NUM
esrj-89750	25	17	to	to	PART
esrj-89750	25	18	make	make	VERB
esrj-89750	25	19	better	well	ADJ
esrj-89750	25	20	use	use	NOUN
esrj-89750	25	21	of	of	ADP
esrj-89750	25	22	land	land	NOUN
esrj-89750	25	23	resources	resource	NOUN
esrj-89750	25	24	and	and	CCONJ
esrj-89750	25	25	develop	develop	VERB
esrj-89750	25	26	agriculture	agriculture	NOUN
esrj-89750	25	27	in	in	ADP
esrj-89750	25	28	china	china	PROPN
esrj-89750	25	29	.	.	PUNCT
esrj-89750	26	1	the	the	DET
esrj-89750	26	2	survey	survey	NOUN
esrj-89750	26	3	contents	content	NOUN
esrj-89750	26	4	generally	generally	ADV
esrj-89750	26	5	include	include	VERB
esrj-89750	26	6	soil	soil	NOUN
esrj-89750	26	7	formation	formation	NOUN
esrj-89750	26	8	factors	factor	NOUN
esrj-89750	26	9	,	,	PUNCT
esrj-89750	26	10	description	description	NOUN
esrj-89750	26	11	of	of	ADP
esrj-89750	26	12	typical	typical	ADJ
esrj-89750	26	13	soil	soil	NOUN
esrj-89750	26	14	profiles	profile	NOUN
esrj-89750	26	15	,	,	PUNCT
esrj-89750	26	16	classification	classification	NOUN
esrj-89750	26	17	of	of	ADP
esrj-89750	26	18	soil	soil	NOUN
esrj-89750	26	19	types	type	NOUN
esrj-89750	26	20	,	,	PUNCT
esrj-89750	26	21	determination	determination	NOUN
esrj-89750	26	22	of	of	ADP
esrj-89750	26	23	soil	soil	NOUN
esrj-89750	26	24	physical	physical	ADJ
esrj-89750	26	25	and	and	CCONJ
esrj-89750	26	26	chemical	chemical	NOUN
esrj-89750	26	27	properties	property	NOUN
esrj-89750	26	28	,	,	PUNCT
esrj-89750	26	29	soil	soil	NOUN
esrj-89750	26	30	evaluation	evaluation	NOUN
esrj-89750	26	31	,	,	PUNCT
esrj-89750	26	32	and	and	CCONJ
esrj-89750	26	33	low	low	ADJ
esrj-89750	26	34	-	-	PUNCT
esrj-89750	26	35	yield	yield	NOUN
esrj-89750	26	36	soil	soil	NOUN
esrj-89750	26	37	improvement	improvement	NOUN
esrj-89750	26	38	planning	planning	NOUN
esrj-89750	26	39	.	.	PUNCT
esrj-89750	27	1	however	however	ADV
esrj-89750	27	2	,	,	PUNCT
esrj-89750	27	3	in	in	ADP
esrj-89750	27	4	the	the	DET
esrj-89750	27	5	face	face	NOUN
esrj-89750	27	6	of	of	ADP
esrj-89750	27	7	such	such	ADJ
esrj-89750	27	8	substantial	substantial	ADJ
esrj-89750	27	9	land	land	NOUN
esrj-89750	27	10	resources	resource	NOUN
esrj-89750	27	11	and	and	CCONJ
esrj-89750	27	12	continuous	continuous	ADJ
esrj-89750	27	13	changes	change	NOUN
esrj-89750	27	14	,	,	PUNCT
esrj-89750	27	15	the	the	DET
esrj-89750	27	16	previous	previous	ADJ
esrj-89750	27	17	two	two	NUM
esrj-89750	27	18	soil	soil	NOUN
esrj-89750	27	19	censuses	census	NOUN
esrj-89750	27	20	have	have	AUX
esrj-89750	27	21	become	become	VERB
esrj-89750	27	22	very	very	ADV
esrj-89750	27	23	difficult	difficult	ADJ
esrj-89750	27	24	,	,	PUNCT
esrj-89750	27	25	especially	especially	ADV
esrj-89750	27	26	in	in	ADP
esrj-89750	27	27	the	the	DET
esrj-89750	27	28	classification	classification	NOUN
esrj-89750	27	29	process	process	NOUN
esrj-89750	27	30	,	,	PUNCT
esrj-89750	27	31	it	it	PRON
esrj-89750	27	32	spent	spend	VERB
esrj-89750	27	33	a	a	DET
esrj-89750	27	34	lot	lot	NOUN
esrj-89750	27	35	of	of	ADP
esrj-89750	27	36	time	time	NOUN
esrj-89750	27	37	,	,	PUNCT
esrj-89750	27	38	material	material	NOUN
esrj-89750	27	39	,	,	PUNCT
esrj-89750	27	40	and	and	CCONJ
esrj-89750	27	41	financial	financial	ADJ
esrj-89750	27	42	resources	resource	NOUN
esrj-89750	27	43	.	.	PUNCT
esrj-89750	28	1	with	with	ADP
esrj-89750	28	2	the	the	DET
esrj-89750	28	3	emergence	emergence	NOUN
esrj-89750	28	4	and	and	CCONJ
esrj-89750	28	5	development	development	NOUN
esrj-89750	28	6	of	of	ADP
esrj-89750	28	7	remote	remote	ADJ
esrj-89750	28	8	sensing	sense	VERB
esrj-89750	28	9	technology	technology	NOUN
esrj-89750	28	10	,	,	PUNCT
esrj-89750	28	11	the	the	DET
esrj-89750	28	12	work	work	NOUN
esrj-89750	28	13	of	of	ADP
esrj-89750	28	14	soil	soil	NOUN
esrj-89750	28	15	survey	survey	NOUN
esrj-89750	28	16	has	have	AUX
esrj-89750	28	17	become	become	VERB
esrj-89750	28	18	simple	simple	ADJ
esrj-89750	28	19	.	.	PUNCT
esrj-89750	29	1	however	however	ADV
esrj-89750	29	2	,	,	PUNCT
esrj-89750	29	3	with	with	ADP
esrj-89750	29	4	the	the	DET
esrj-89750	29	5	improvement	improvement	NOUN
esrj-89750	29	6	of	of	ADP
esrj-89750	29	7	remote	remote	ADJ
esrj-89750	29	8	sensing	sense	VERB
esrj-89750	29	9	image	image	NOUN
esrj-89750	29	10	resolution	resolution	NOUN
esrj-89750	29	11	and	and	CCONJ
esrj-89750	29	12	the	the	DET
esrj-89750	29	13	increase	increase	NOUN
esrj-89750	29	14	of	of	ADP
esrj-89750	29	15	data	datum	NOUN
esrj-89750	29	16	volume	volume	NOUN
esrj-89750	29	17	,	,	PUNCT
esrj-89750	29	18	the	the	DET
esrj-89750	29	19	collected	collect	VERB
esrj-89750	29	20	soil	soil	NOUN
esrj-89750	29	21	images	image	NOUN
esrj-89750	29	22	are	be	AUX
esrj-89750	29	23	more	more	ADV
esrj-89750	29	24	diverse	diverse	ADJ
esrj-89750	29	25	and	and	CCONJ
esrj-89750	29	26	more	more	ADV
esrj-89750	29	27	productive	productive	ADJ
esrj-89750	29	28	,	,	PUNCT
esrj-89750	29	29	but	but	CCONJ
esrj-89750	29	30	to	to	ADP
esrj-89750	29	31	a	a	DET
esrj-89750	29	32	certain	certain	ADJ
esrj-89750	29	33	extent	extent	NOUN
esrj-89750	29	34	,	,	PUNCT
esrj-89750	29	35	it	it	PRON
esrj-89750	29	36	increases	increase	VERB
esrj-89750	29	37	the	the	DET
esrj-89750	29	38	difficulty	difficulty	NOUN
esrj-89750	29	39	of	of	ADP
esrj-89750	29	40	classification	classification	NOUN
esrj-89750	29	41	(	(	PUNCT
esrj-89750	29	42	chen	chen	PROPN
esrj-89750	29	43	et	et	PROPN
esrj-89750	29	44	al	al	PROPN
esrj-89750	29	45	.	.	PROPN
esrj-89750	29	46	,	,	PUNCT
esrj-89750	29	47	2019	2019	NUM
esrj-89750	29	48	)	)	PUNCT
esrj-89750	29	49	.	.	PUNCT
esrj-89750	30	1	under	under	ADP
esrj-89750	30	2	the	the	DET
esrj-89750	30	3	above	above	ADJ
esrj-89750	30	4	background	background	NOUN
esrj-89750	30	5	,	,	PUNCT
esrj-89750	30	6	relevant	relevant	ADJ
esrj-89750	30	7	scholars	scholar	NOUN
esrj-89750	30	8	at	at	ADP
esrj-89750	30	9	home	home	ADV
esrj-89750	30	10	and	and	CCONJ
esrj-89750	30	11	abroad	abroad	ADV
esrj-89750	30	12	have	have	AUX
esrj-89750	30	13	conducted	conduct	VERB
esrj-89750	30	14	in	in	ADP
esrj-89750	30	15	-	-	PUNCT
esrj-89750	30	16	depth	depth	NOUN
esrj-89750	30	17	research	research	NOUN
esrj-89750	30	18	on	on	ADP
esrj-89750	30	19	soil	soil	NOUN
esrj-89750	30	20	remote	remote	ADJ
esrj-89750	30	21	sensing	sense	VERB
esrj-89750	30	22	image	image	NOUN
esrj-89750	30	23	classification	classification	NOUN
esrj-89750	30	24	and	and	CCONJ
esrj-89750	30	25	proposed	propose	VERB
esrj-89750	30	26	many	many	ADJ
esrj-89750	30	27	methods	method	NOUN
esrj-89750	30	28	,	,	PUNCT
esrj-89750	30	29	such	such	ADJ
esrj-89750	30	30	as	as	ADP
esrj-89750	30	31	soil	soil	NOUN
esrj-89750	30	32	remote	remote	ADJ
esrj-89750	30	33	sensing	sense	VERB
esrj-89750	30	34	image	image	NOUN
esrj-89750	30	35	classification	classification	NOUN
esrj-89750	30	36	methods	method	NOUN
esrj-89750	30	37	,	,	PUNCT
esrj-89750	30	38	based	base	VERB
esrj-89750	30	39	on	on	ADP
esrj-89750	30	40	deep	deep	ADJ
esrj-89750	30	41	learning	learning	NOUN
esrj-89750	30	42	.	.	PUNCT
esrj-89750	31	1	it	it	PRON
esrj-89750	31	2	is	be	AUX
esrj-89750	31	3	based	base	VERB
esrj-89750	31	4	on	on	ADP
esrj-89750	31	5	artificial	artificial	ADJ
esrj-89750	31	6	neural	neural	ADJ
esrj-89750	31	7	network	network	NOUN
esrj-89750	31	8	architecture	architecture	NOUN
esrj-89750	31	9	,	,	PUNCT
esrj-89750	31	10	such	such	ADJ
esrj-89750	31	11	as	as	ADP
esrj-89750	31	12	convolution	convolution	NOUN
esrj-89750	31	13	neural	neural	ADJ
esrj-89750	31	14	network	network	NOUN
esrj-89750	31	15	,	,	PUNCT
esrj-89750	31	16	deep	deep	ADJ
esrj-89750	31	17	neural	neural	ADJ
esrj-89750	31	18	network	network	NOUN
esrj-89750	31	19	,	,	PUNCT
esrj-89750	31	20	deep	deep	ADJ
esrj-89750	31	21	confidence	confidence	NOUN
esrj-89750	31	22	network	network	NOUN
esrj-89750	31	23	,	,	PUNCT
esrj-89750	31	24	to	to	PART
esrj-89750	31	25	learn	learn	VERB
esrj-89750	31	26	more	more	ADV
esrj-89750	31	27	useful	useful	ADJ
esrj-89750	31	28	features	feature	NOUN
esrj-89750	31	29	,	,	PUNCT
esrj-89750	31	30	thus	thus	ADV
esrj-89750	31	31	ultimately	ultimately	ADV
esrj-89750	31	32	realizing	realize	VERB
esrj-89750	31	33	classification	classification	NOUN
esrj-89750	31	34	.	.	PUNCT
esrj-89750	32	1	the	the	DET
esrj-89750	32	2	advantage	advantage	NOUN
esrj-89750	32	3	of	of	ADP
esrj-89750	32	4	this	this	DET
esrj-89750	32	5	method	method	NOUN
esrj-89750	32	6	is	be	AUX
esrj-89750	32	7	that	that	SCONJ
esrj-89750	32	8	it	it	PRON
esrj-89750	32	9	has	have	VERB
esrj-89750	32	10	better	well	ADJ
esrj-89750	32	11	transfer	transfer	VERB
esrj-89750	32	12	learning	learn	VERB
esrj-89750	32	13	property	property	NOUN
esrj-89750	32	14	.	.	PUNCT
esrj-89750	33	1	the	the	DET
esrj-89750	33	2	disadvantage	disadvantage	NOUN
esrj-89750	33	3	is	be	AUX
esrj-89750	33	4	that	that	SCONJ
esrj-89750	33	5	model	model	NOUN
esrj-89750	33	6	validation	validation	NOUN
esrj-89750	33	7	is	be	AUX
esrj-89750	33	8	complicated	complicated	ADJ
esrj-89750	33	9	and	and	CCONJ
esrj-89750	33	10	cumbersome	cumbersome	ADJ
esrj-89750	33	11	.	.	PUNCT
esrj-89750	34	1	for	for	ADP
esrj-89750	34	2	example	example	NOUN
esrj-89750	34	3	,	,	PUNCT
esrj-89750	34	4	the	the	DET
esrj-89750	34	5	principle	principle	NOUN
esrj-89750	34	6	of	of	ADP
esrj-89750	34	7	a	a	DET
esrj-89750	34	8	classification	classification	NOUN
esrj-89750	34	9	method	method	NOUN
esrj-89750	34	10	for	for	ADP
esrj-89750	34	11	soil	soil	NOUN
esrj-89750	34	12	remote	remote	ADJ
esrj-89750	34	13	sensing	sense	VERB
esrj-89750	34	14	image	image	NOUN
esrj-89750	34	15	based	base	VERB
esrj-89750	34	16	on	on	ADP
esrj-89750	34	17	maximum	maximum	ADJ
esrj-89750	34	18	likelihood	likelihood	NOUN
esrj-89750	34	19	estimation	estimation	NOUN
esrj-89750	34	20	is	be	AUX
esrj-89750	34	21	based	base	VERB
esrj-89750	34	22	on	on	ADP
esrj-89750	34	23	calculating	calculate	VERB
esrj-89750	34	24	the	the	DET
esrj-89750	34	25	probability	probability	NOUN
esrj-89750	34	26	that	that	SCONJ
esrj-89750	34	27	a	a	DET
esrj-89750	34	28	pixel	pixel	NOUN
esrj-89750	34	29	belongs	belong	VERB
esrj-89750	34	30	to	to	ADP
esrj-89750	34	31	each	each	DET
esrj-89750	34	32	class	class	NOUN
esrj-89750	34	33	in	in	ADP
esrj-89750	34	34	a	a	DET
esrj-89750	34	35	pre	pre	ADJ
esrj-89750	34	36	-	-	ADJ
esrj-89750	34	37	set	set	ADJ
esrj-89750	34	38	m	m	NOUN
esrj-89750	34	39	-	-	PUNCT
esrj-89750	34	40	class	class	NOUN
esrj-89750	34	41	data	datum	NOUN
esrj-89750	34	42	set	set	NOUN
esrj-89750	34	43	,	,	PUNCT
esrj-89750	34	44	and	and	CCONJ
esrj-89750	34	45	then	then	ADV
esrj-89750	34	46	dividing	divide	VERB
esrj-89750	34	47	it	it	PRON
esrj-89750	34	48	into	into	ADP
esrj-89750	34	49	the	the	DET
esrj-89750	34	50	most	most	ADV
esrj-89750	34	51	probabilistic	probabilistic	ADJ
esrj-89750	34	52	class	class	NOUN
esrj-89750	34	53	.	.	PUNCT
esrj-89750	35	1	the	the	DET
esrj-89750	35	2	advantage	advantage	NOUN
esrj-89750	35	3	of	of	ADP
esrj-89750	35	4	this	this	DET
esrj-89750	35	5	method	method	NOUN
esrj-89750	35	6	is	be	AUX
esrj-89750	35	7	that	that	SCONJ
esrj-89750	35	8	it	it	PRON
esrj-89750	35	9	has	have	VERB
esrj-89750	35	10	evident	evident	ADJ
esrj-89750	35	11	parameter	parameter	NOUN
esrj-89750	35	12	interpretation	interpretation	NOUN
esrj-89750	35	13	ability	ability	NOUN
esrj-89750	35	14	,	,	PUNCT
esrj-89750	35	15	and	and	CCONJ
esrj-89750	35	16	is	be	AUX
esrj-89750	35	17	easy	easy	ADJ
esrj-89750	35	18	to	to	PART
esrj-89750	35	19	fuse	fuse	VERB
esrj-89750	35	20	with	with	ADP
esrj-89750	35	21	prior	prior	ADJ
esrj-89750	35	22	knowledge	knowledge	NOUN
esrj-89750	35	23	.	.	PUNCT
esrj-89750	36	1	the	the	DET
esrj-89750	36	2	algorithm	algorithm	NOUN
esrj-89750	36	3	is	be	AUX
esrj-89750	36	4	simple	simple	ADJ
esrj-89750	36	5	and	and	CCONJ
esrj-89750	36	6	easy	easy	ADJ
esrj-89750	36	7	to	to	PART
esrj-89750	36	8	implement	implement	VERB
esrj-89750	36	9	.	.	PUNCT
esrj-89750	37	1	the	the	DET
esrj-89750	37	2	disadvantage	disadvantage	NOUN
esrj-89750	37	3	is	be	AUX
esrj-89750	37	4	that	that	SCONJ
esrj-89750	37	5	it	it	PRON
esrj-89750	37	6	is	be	AUX
esrj-89750	37	7	vulnerable	vulnerable	ADJ
esrj-89750	37	8	to	to	ADP
esrj-89750	37	9	the	the	DET
esrj-89750	37	10	distribution	distribution	NOUN
esrj-89750	37	11	of	of	ADP
esrj-89750	37	12	categories	category	NOUN
esrj-89750	37	13	in	in	ADP
esrj-89750	37	14	feature	feature	NOUN
esrj-89750	37	15	space	space	NOUN
esrj-89750	37	16	and	and	CCONJ
esrj-89750	37	17	the	the	DET
esrj-89750	37	18	selection	selection	NOUN
esrj-89750	37	19	of	of	ADP
esrj-89750	37	20	samples	sample	NOUN
esrj-89750	37	21	.	.	PUNCT
esrj-89750	38	1	once	once	SCONJ
esrj-89750	38	2	the	the	DET
esrj-89750	38	3	distribution	distribution	NOUN
esrj-89750	38	4	is	be	AUX
esrj-89750	38	5	discrete	discrete	ADJ
esrj-89750	38	6	,	,	PUNCT
esrj-89750	38	7	or	or	CCONJ
esrj-89750	38	8	the	the	DET
esrj-89750	38	9	selected	select	VERB
esrj-89750	38	10	samples	sample	NOUN
esrj-89750	38	11	are	be	AUX
esrj-89750	38	12	not	not	PART
esrj-89750	38	13	representative	representative	ADJ
esrj-89750	38	14	,	,	PUNCT
esrj-89750	38	15	the	the	DET
esrj-89750	38	16	classification	classification	NOUN
esrj-89750	38	17	results	result	NOUN
esrj-89750	38	18	will	will	AUX
esrj-89750	38	19	deviate	deviate	VERB
esrj-89750	38	20	significantly	significantly	ADV
esrj-89750	38	21	from	from	ADP
esrj-89750	38	22	the	the	DET
esrj-89750	38	23	actual	actual	ADJ
esrj-89750	38	24	situation	situation	NOUN
esrj-89750	38	25	(	(	PUNCT
esrj-89750	38	26	cheng	cheng	PROPN
esrj-89750	38	27	et	et	PROPN
esrj-89750	38	28	al	al	PROPN
esrj-89750	38	29	.	.	PROPN
esrj-89750	38	30	,	,	PUNCT
esrj-89750	38	31	2018	2018	NUM
esrj-89750	38	32	)	)	PUNCT
esrj-89750	38	33	.	.	PUNCT
esrj-89750	39	1	given	give	VERB
esrj-89750	39	2	the	the	DET
esrj-89750	39	3	above	above	ADJ
esrj-89750	39	4	situation	situation	NOUN
esrj-89750	39	5	,	,	PUNCT
esrj-89750	39	6	this	this	DET
esrj-89750	39	7	paper	paper	NOUN
esrj-89750	39	8	combines	combine	VERB
esrj-89750	39	9	deep	deep	ADJ
esrj-89750	39	10	learning	learning	NOUN
esrj-89750	39	11	with	with	ADP
esrj-89750	39	12	maximum	maximum	ADJ
esrj-89750	39	13	likelihood	likelihood	NOUN
esrj-89750	39	14	estimation	estimation	NOUN
esrj-89750	39	15	and	and	CCONJ
esrj-89750	39	16	proposes	propose	VERB
esrj-89750	39	17	a	a	DET
esrj-89750	39	18	maximum	maximum	ADJ
esrj-89750	39	19	likelihood	likelihood	NOUN
esrj-89750	39	20	classification	classification	NOUN
esrj-89750	39	21	method	method	NOUN
esrj-89750	39	22	for	for	ADP
esrj-89750	39	23	soil	soil	NOUN
esrj-89750	39	24	remote	remote	ADJ
esrj-89750	39	25	sensing	sensing	NOUN
esrj-89750	39	26	images	image	NOUN
esrj-89750	39	27	based	base	VERB
esrj-89750	39	28	on	on	ADP
esrj-89750	39	29	deep	deep	ADJ
esrj-89750	39	30	learning	learning	NOUN
esrj-89750	39	31	.	.	PUNCT
esrj-89750	40	1	the	the	DET
esrj-89750	40	2	method	method	NOUN
esrj-89750	40	3	is	be	AUX
esrj-89750	40	4	divided	divide	VERB
esrj-89750	40	5	into	into	ADP
esrj-89750	40	6	two	two	NUM
esrj-89750	40	7	parts	part	NOUN
esrj-89750	40	8	.	.	PUNCT
esrj-89750	41	1	the	the	DET
esrj-89750	41	2	first	first	ADJ
esrj-89750	41	3	part	part	NOUN
esrj-89750	41	4	is	be	AUX
esrj-89750	41	5	to	to	PART
esrj-89750	41	6	detect	detect	VERB
esrj-89750	41	7	remote	remote	ADJ
esrj-89750	41	8	sensing	sense	VERB
esrj-89750	41	9	image	image	NOUN
esrj-89750	41	10	targets	target	NOUN
esrj-89750	41	11	by	by	ADP
esrj-89750	41	12	deep	deep	ADJ
esrj-89750	41	13	learning	learning	NOUN
esrj-89750	41	14	and	and	CCONJ
esrj-89750	41	15	extracts	extract	VERB
esrj-89750	41	16	classification	classification	NOUN
esrj-89750	41	17	features	feature	NOUN
esrj-89750	41	18	of	of	ADP
esrj-89750	41	19	various	various	ADJ
esrj-89750	41	20	image	image	NOUN
esrj-89750	41	21	targets	target	NOUN
esrj-89750	41	22	.	.	PUNCT
esrj-89750	42	1	the	the	DET
esrj-89750	42	2	second	second	ADJ
esrj-89750	42	3	part	part	NOUN
esrj-89750	42	4	is	be	AUX
esrj-89750	42	5	to	to	PART
esrj-89750	42	6	classify	classify	VERB
esrj-89750	42	7	soil	soil	NOUN
esrj-89750	42	8	remote	remote	ADJ
esrj-89750	42	9	sensing	sensing	NOUN
esrj-89750	42	10	images	image	NOUN
esrj-89750	42	11	by	by	ADP
esrj-89750	42	12	maximum	maximum	ADJ
esrj-89750	42	13	likelihood	likelihood	NOUN
esrj-89750	42	14	estimation	estimation	NOUN
esrj-89750	42	15	algorithm	algorithm	NOUN
esrj-89750	42	16	based	base	VERB
esrj-89750	42	17	on	on	ADP
esrj-89750	42	18	the	the	DET
esrj-89750	42	19	first	first	ADJ
esrj-89750	42	20	part	part	NOUN
esrj-89750	42	21	and	and	CCONJ
esrj-89750	42	22	produce	produce	VERB
esrj-89750	42	23	soil	soil	NOUN
esrj-89750	42	24	-	-	PUNCT
esrj-89750	42	25	type	type	NOUN
esrj-89750	42	26	maps	map	NOUN
esrj-89750	42	27	(	(	PUNCT
esrj-89750	42	28	demattê	demattê	NOUN
esrj-89750	42	29	et	et	PROPN
esrj-89750	42	30	al	al	PROPN
esrj-89750	42	31	.	.	PROPN
esrj-89750	42	32	,	,	PUNCT
esrj-89750	42	33	2017	2017	NUM
esrj-89750	42	34	)	)	PUNCT
esrj-89750	42	35	.	.	PUNCT
esrj-89750	43	1	in	in	ADP
esrj-89750	43	2	the	the	DET
esrj-89750	43	3	experimental	experimental	ADJ
esrj-89750	43	4	part	part	NOUN
esrj-89750	43	5	,	,	PUNCT
esrj-89750	43	6	a	a	DET
esrj-89750	43	7	remote	remote	ADJ
esrj-89750	43	8	sensing	sense	VERB
esrj-89750	43	9	image	image	NOUN
esrj-89750	43	10	of	of	ADP
esrj-89750	43	11	a	a	DET
esrj-89750	43	12	region	region	NOUN
esrj-89750	43	13	is	be	AUX
esrj-89750	43	14	taken	take	VERB
esrj-89750	43	15	as	as	ADP
esrj-89750	43	16	an	an	DET
esrj-89750	43	17	example	example	NOUN
esrj-89750	43	18	to	to	PART
esrj-89750	43	19	test	test	VERB
esrj-89750	43	20	the	the	DET
esrj-89750	43	21	classification	classification	NOUN
esrj-89750	43	22	performance	performance	NOUN
esrj-89750	43	23	.	.	PUNCT
esrj-89750	44	1	by	by	ADP
esrj-89750	44	2	drawing	draw	VERB
esrj-89750	44	3	roc	roc	PROPN
esrj-89750	44	4	curves	curve	NOUN
esrj-89750	44	5	,	,	PUNCT
esrj-89750	44	6	it	it	PRON
esrj-89750	44	7	is	be	AUX
esrj-89750	44	8	concluded	conclude	VERB
esrj-89750	44	9	that	that	SCONJ
esrj-89750	44	10	the	the	DET
esrj-89750	44	11	combined	combine	VERB
esrj-89750	44	12	soil	soil	NOUN
esrj-89750	44	13	remote	remote	ADJ
esrj-89750	44	14	sensing	sense	VERB
esrj-89750	44	15	image	image	NOUN
esrj-89750	44	16	has	have	VERB
esrj-89750	44	17	better	well	ADJ
esrj-89750	44	18	classification	classification	NOUN
esrj-89750	44	19	performance	performance	NOUN
esrj-89750	44	20	.	.	PUNCT
esrj-89750	45	1	this	this	DET
esrj-89750	45	2	improvement	improvement	NOUN
esrj-89750	45	3	provides	provide	VERB
esrj-89750	45	4	a	a	DET
esrj-89750	45	5	new	new	ADJ
esrj-89750	45	6	idea	idea	NOUN
esrj-89750	45	7	for	for	ADP
esrj-89750	45	8	soil	soil	NOUN
esrj-89750	45	9	classification	classification	NOUN
esrj-89750	45	10	and	and	CCONJ
esrj-89750	45	11	mapping	mapping	NOUN
esrj-89750	45	12	of	of	ADP
esrj-89750	45	13	current	current	ADJ
esrj-89750	45	14	remote	remote	ADJ
esrj-89750	45	15	sensing	sense	VERB
esrj-89750	45	16	data	datum	NOUN
esrj-89750	45	17	,	,	PUNCT
esrj-89750	45	18	facilitates	facilitate	VERB
esrj-89750	45	19	soil	soil	NOUN
esrj-89750	45	20	census	census	NOUN
esrj-89750	45	21	,	,	PUNCT
esrj-89750	45	22	improves	improve	VERB
esrj-89750	45	23	land	land	NOUN
esrj-89750	45	24	use	use	NOUN
esrj-89750	45	25	rate	rate	NOUN
esrj-89750	45	26	to	to	ADP
esrj-89750	45	27	a	a	DET
esrj-89750	45	28	certain	certain	ADJ
esrj-89750	45	29	extent	extent	NOUN
esrj-89750	45	30	,	,	PUNCT
esrj-89750	45	31	and	and	CCONJ
esrj-89750	45	32	promotes	promote	VERB
esrj-89750	45	33	agriculture	agriculture	NOUN
esrj-89750	45	34	,	,	PUNCT
esrj-89750	45	35	animal	animal	NOUN
esrj-89750	45	36	husbandry	husbandry	NOUN
esrj-89750	45	37	,	,	PUNCT
esrj-89750	45	38	and	and	CCONJ
esrj-89750	45	39	forestry	forestry	NOUN
esrj-89750	45	40	in	in	ADP
esrj-89750	45	41	china	china	PROPN
esrj-89750	45	42	.	.	PUNCT
esrj-89750	46	1	maximum	maximum	ADJ
esrj-89750	46	2	likelihood	likelihood	NOUN
esrj-89750	46	3	classification	classification	NOUN
esrj-89750	46	4	of	of	ADP
esrj-89750	46	5	soil	soil	NOUN
esrj-89750	46	6	remote	remote	ADJ
esrj-89750	46	7	sensing	sense	VERB
esrj-89750	46	8	image	image	NOUN
esrj-89750	46	9	based	base	VERB
esrj-89750	46	10	on	on	ADP
esrj-89750	46	11	deep	deep	ADJ
esrj-89750	46	12	learning	learn	VERB
esrj-89750	46	13	soil	soil	NOUN
esrj-89750	46	14	remote	remote	ADJ
esrj-89750	46	15	sensing	sense	VERB
esrj-89750	46	16	image	image	NOUN
esrj-89750	46	17	is	be	AUX
esrj-89750	46	18	a	a	DET
esrj-89750	46	19	kind	kind	NOUN
esrj-89750	46	20	of	of	ADP
esrj-89750	46	21	soil	soil	NOUN
esrj-89750	46	22	remote	remote	ADJ
esrj-89750	46	23	sensing	sense	VERB
esrj-89750	46	24	image	image	NOUN
esrj-89750	46	25	obtained	obtain	VERB
esrj-89750	46	26	by	by	ADP
esrj-89750	46	27	various	various	ADJ
esrj-89750	46	28	sensing	sense	VERB
esrj-89750	46	29	devices	device	NOUN
esrj-89750	46	30	(	(	PUNCT
esrj-89750	46	31	such	such	ADJ
esrj-89750	46	32	as	as	ADP
esrj-89750	46	33	radar	radar	NOUN
esrj-89750	46	34	,	,	PUNCT
esrj-89750	46	35	camera	camera	NOUN
esrj-89750	46	36	,	,	PUNCT
esrj-89750	46	37	scanner	scanner	NOUN
esrj-89750	46	38	,	,	PUNCT
esrj-89750	46	39	etc	etc	X
esrj-89750	46	40	.	.	X
esrj-89750	46	41	)	)	PUNCT
esrj-89750	46	42	.	.	PUNCT
esrj-89750	47	1	its	its	PRON
esrj-89750	47	2	acquisition	acquisition	NOUN
esrj-89750	47	3	principle	principle	NOUN
esrj-89750	47	4	process	process	NOUN
esrj-89750	47	5	is	be	AUX
esrj-89750	47	6	shown	show	VERB
esrj-89750	47	7	in	in	ADP
esrj-89750	47	8	figure	figure	NOUN
esrj-89750	47	9	1	1	NUM
esrj-89750	47	10	,	,	PUNCT
esrj-89750	47	11	and	and	CCONJ
esrj-89750	47	12	the	the	DET
esrj-89750	47	13	form	form	NOUN
esrj-89750	47	14	of	of	ADP
esrj-89750	47	15	the	the	DET
esrj-89750	47	16	obtained	obtain	VERB
esrj-89750	47	17	remote	remote	ADJ
esrj-89750	47	18	sensing	sense	VERB
esrj-89750	47	19	image	image	NOUN
esrj-89750	47	20	is	be	AUX
esrj-89750	47	21	shown	show	VERB
esrj-89750	47	22	in	in	ADP
esrj-89750	47	23	figure	figure	NOUN
esrj-89750	47	24	2	2	NUM
esrj-89750	47	25	.	.	PUNCT
esrj-89750	48	1	the	the	DET
esrj-89750	48	2	obtained	obtain	VERB
esrj-89750	48	3	soil	soil	NOUN
esrj-89750	48	4	remote	remote	ADJ
esrj-89750	48	5	sensing	sensing	NOUN
esrj-89750	48	6	images	image	NOUN
esrj-89750	48	7	can	can	AUX
esrj-89750	48	8	accurately	accurately	ADV
esrj-89750	48	9	identify	identify	VERB
esrj-89750	48	10	and	and	CCONJ
esrj-89750	48	11	classify	classify	VERB
esrj-89750	48	12	soil	soil	NOUN
esrj-89750	48	13	types	type	NOUN
esrj-89750	48	14	,	,	PUNCT
esrj-89750	48	15	make	make	VERB
esrj-89750	48	16	soil	soil	NOUN
esrj-89750	48	17	maps	map	NOUN
esrj-89750	48	18	,	,	PUNCT
esrj-89750	48	19	and	and	CCONJ
esrj-89750	48	20	analyze	analyze	VERB
esrj-89750	48	21	soil	soil	NOUN
esrj-89750	48	22	distribution	distribution	NOUN
esrj-89750	48	23	law	law	NOUN
esrj-89750	48	24	,	,	PUNCT
esrj-89750	48	25	which	which	PRON
esrj-89750	48	26	brings	bring	VERB
esrj-89750	48	27	great	great	ADJ
esrj-89750	48	28	convenience	convenience	NOUN
esrj-89750	48	29	to	to	PART
esrj-89750	48	30	soil	soil	NOUN
esrj-89750	48	31	survey	survey	NOUN
esrj-89750	48	32	.	.	PUNCT
esrj-89750	49	1	to	to	PART
esrj-89750	49	2	improve	improve	VERB
esrj-89750	49	3	the	the	DET
esrj-89750	49	4	classification	classification	NOUN
esrj-89750	49	5	quality	quality	NOUN
esrj-89750	49	6	of	of	ADP
esrj-89750	49	7	soil	soil	NOUN
esrj-89750	49	8	remote	remote	ADJ
esrj-89750	49	9	sensing	sensing	NOUN
esrj-89750	49	10	images	image	NOUN
esrj-89750	49	11	,	,	PUNCT
esrj-89750	49	12	this	this	DET
esrj-89750	49	13	paper	paper	NOUN
esrj-89750	49	14	studies	study	NOUN
esrj-89750	49	15	a	a	DET
esrj-89750	49	16	more	more	ADV
esrj-89750	49	17	effective	effective	ADJ
esrj-89750	49	18	classification	classification	NOUN
esrj-89750	49	19	method	method	NOUN
esrj-89750	49	20	,	,	PUNCT
esrj-89750	49	21	which	which	PRON
esrj-89750	49	22	combines	combine	VERB
esrj-89750	49	23	deep	deep	ADJ
esrj-89750	49	24	learning	learning	NOUN
esrj-89750	49	25	with	with	ADP
esrj-89750	49	26	maximum	maximum	ADJ
esrj-89750	49	27	likelihood	likelihood	NOUN
esrj-89750	49	28	estimation	estimation	NOUN
esrj-89750	49	29	to	to	PART
esrj-89750	49	30	make	make	VERB
esrj-89750	49	31	up	up	ADP
esrj-89750	49	32	for	for	ADP
esrj-89750	49	33	each	each	DET
esrj-89750	49	34	other	other	ADJ
esrj-89750	49	35	’s	’s	PART
esrj-89750	49	36	shortcomings	shortcoming	NOUN
esrj-89750	49	37	(	(	PUNCT
esrj-89750	49	38	fitak	fitak	PROPN
esrj-89750	49	39	&	&	CCONJ
esrj-89750	49	40	johnsen	johnsen	PROPN
esrj-89750	49	41	,	,	PUNCT
esrj-89750	49	42	2017	2017	NUM
esrj-89750	49	43	)	)	PUNCT
esrj-89750	49	44	.	.	PUNCT
esrj-89750	50	1	the	the	DET
esrj-89750	50	2	research	research	NOUN
esrj-89750	50	3	on	on	ADP
esrj-89750	50	4	maximum	maximum	ADJ
esrj-89750	50	5	likelihood	likelihood	NOUN
esrj-89750	50	6	classification	classification	NOUN
esrj-89750	50	7	of	of	ADP
esrj-89750	50	8	soil	soil	NOUN
esrj-89750	50	9	remote	remote	ADJ
esrj-89750	50	10	sensing	sensing	NOUN
esrj-89750	50	11	images	image	NOUN
esrj-89750	50	12	based	base	VERB
esrj-89750	50	13	on	on	ADP
esrj-89750	50	14	deep	deep	ADJ
esrj-89750	50	15	learning	learning	NOUN
esrj-89750	50	16	is	be	AUX
esrj-89750	50	17	mainly	mainly	ADV
esrj-89750	50	18	divided	divide	VERB
esrj-89750	50	19	into	into	ADP
esrj-89750	50	20	four	four	NUM
esrj-89750	50	21	stages	stage	NOUN
esrj-89750	50	22	:	:	PUNCT
esrj-89750	50	23	the	the	DET
esrj-89750	50	24	first	first	ADJ
esrj-89750	50	25	stage	stage	NOUN
esrj-89750	50	26	is	be	AUX
esrj-89750	50	27	to	to	PART
esrj-89750	50	28	preprocess	preprocess	VERB
esrj-89750	50	29	the	the	DET
esrj-89750	50	30	soil	soil	NOUN
esrj-89750	50	31	remote	remote	ADJ
esrj-89750	50	32	sensing	sensing	NOUN
esrj-89750	50	33	images	image	NOUN
esrj-89750	50	34	acquired	acquire	VERB
esrj-89750	50	35	by	by	ADP
esrj-89750	50	36	remote	remote	ADJ
esrj-89750	50	37	sensing	sense	VERB
esrj-89750	50	38	equipment	equipment	NOUN
esrj-89750	50	39	;	;	PUNCT
esrj-89750	50	40	the	the	DET
esrj-89750	50	41	second	second	ADJ
esrj-89750	50	42	stage	stage	NOUN
esrj-89750	50	43	is	be	AUX
esrj-89750	50	44	to	to	PART
esrj-89750	50	45	detect	detect	VERB
esrj-89750	50	46	the	the	DET
esrj-89750	50	47	targets	target	NOUN
esrj-89750	50	48	of	of	ADP
esrj-89750	50	49	remote	remote	ADJ
esrj-89750	50	50	sensing	sensing	NOUN
esrj-89750	50	51	images	image	NOUN
esrj-89750	50	52	by	by	ADP
esrj-89750	50	53	using	use	VERB
esrj-89750	50	54	deep	deep	ADJ
esrj-89750	50	55	learning	learning	NOUN
esrj-89750	50	56	algorithm	algorithm	NOUN
esrj-89750	50	57	and	and	CCONJ
esrj-89750	50	58	extract	extract	VERB
esrj-89750	50	59	the	the	DET
esrj-89750	50	60	classification	classification	NOUN
esrj-89750	50	61	features	feature	NOUN
esrj-89750	50	62	of	of	ADP
esrj-89750	50	63	various	various	ADJ
esrj-89750	50	64	soil	soil	NOUN
esrj-89750	50	65	images	image	NOUN
esrj-89750	50	66	;	;	PUNCT
esrj-89750	50	67	the	the	DET
esrj-89750	50	68	third	third	ADJ
esrj-89750	50	69	stage	stage	NOUN
esrj-89750	50	70	is	be	AUX
esrj-89750	50	71	to	to	PART
esrj-89750	50	72	classify	classify	VERB
esrj-89750	50	73	the	the	DET
esrj-89750	50	74	soil	soil	NOUN
esrj-89750	50	75	remote	remote	ADJ
esrj-89750	50	76	sensing	sensing	NOUN
esrj-89750	50	77	images	image	NOUN
esrj-89750	50	78	based	base	VERB
esrj-89750	50	79	on	on	ADP
esrj-89750	50	80	the	the	DET
esrj-89750	50	81	methods	method	NOUN
esrj-89750	50	82	mentioned	mention	VERB
esrj-89750	50	83	above	above	ADV
esrj-89750	50	84	by	by	ADP
esrj-89750	50	85	using	use	VERB
esrj-89750	50	86	maximum	maximum	ADJ
esrj-89750	50	87	likelihood	likelihood	NOUN
esrj-89750	50	88	estimation	estimation	NOUN
esrj-89750	50	89	method	method	NOUN
esrj-89750	50	90	.	.	PUNCT
esrj-89750	51	1	in	in	ADP
esrj-89750	51	2	the	the	DET
esrj-89750	51	3	fourth	fourth	ADJ
esrj-89750	51	4	stage	stage	NOUN
esrj-89750	51	5	,	,	PUNCT
esrj-89750	51	6	the	the	DET
esrj-89750	51	7	performance	performance	NOUN
esrj-89750	51	8	of	of	ADP
esrj-89750	51	9	maximum	maximum	ADJ
esrj-89750	51	10	likelihood	likelihood	NOUN
esrj-89750	51	11	classification	classification	NOUN
esrj-89750	51	12	of	of	ADP
esrj-89750	51	13	soil	soil	NOUN
esrj-89750	51	14	remote	remote	ADJ
esrj-89750	51	15	sensing	sensing	NOUN
esrj-89750	51	16	images	image	NOUN
esrj-89750	51	17	based	base	VERB
esrj-89750	51	18	on	on	ADP
esrj-89750	51	19	deep	deep	ADJ
esrj-89750	51	20	learning	learning	NOUN
esrj-89750	51	21	is	be	AUX
esrj-89750	51	22	tested	test	VERB
esrj-89750	51	23	by	by	ADP
esrj-89750	51	24	an	an	DET
esrj-89750	51	25	example	example	NOUN
esrj-89750	51	26	to	to	PART
esrj-89750	51	27	ensure	ensure	VERB
esrj-89750	51	28	the	the	DET
esrj-89750	51	29	effectiveness	effectiveness	NOUN
esrj-89750	51	30	and	and	CCONJ
esrj-89750	51	31	practicability	practicability	NOUN
esrj-89750	51	32	of	of	ADP
esrj-89750	51	33	the	the	DET
esrj-89750	51	34	method	method	NOUN
esrj-89750	51	35	(	(	PUNCT
esrj-89750	51	36	handelman	handelman	PROPN
esrj-89750	51	37	&	&	CCONJ
esrj-89750	51	38	chor	chor	PROPN
esrj-89750	51	39	,	,	PUNCT
esrj-89750	51	40	2017	2017	NUM
esrj-89750	51	41	)	)	PUNCT
esrj-89750	51	42	.	.	PUNCT
esrj-89750	52	1	figure	figure	NOUN
esrj-89750	52	2	1	1	NUM
esrj-89750	52	3	.	.	PUNCT
esrj-89750	52	4	principle	principle	NOUN
esrj-89750	52	5	of	of	ADP
esrj-89750	52	6	soil	soil	NOUN
esrj-89750	52	7	remote	remote	ADJ
esrj-89750	52	8	sensing	sense	VERB
esrj-89750	52	9	image	image	NOUN
esrj-89750	52	10	acquisition	acquisition	NOUN
esrj-89750	52	11	figure	figure	NOUN
esrj-89750	52	12	2	2	NUM
esrj-89750	52	13	.	.	PUNCT
esrj-89750	52	14	remote	remote	ADJ
esrj-89750	52	15	sensing	sensing	NOUN
esrj-89750	52	16	images	image	NOUN
esrj-89750	52	17	359maximum	359maximum	PROPN
esrj-89750	52	18	likelihood	likelihood	NOUN
esrj-89750	52	19	classification	classification	NOUN
esrj-89750	52	20	of	of	ADP
esrj-89750	52	21	soil	soil	NOUN
esrj-89750	52	22	remote	remote	ADJ
esrj-89750	52	23	sensing	sense	VERB
esrj-89750	52	24	image	image	NOUN
esrj-89750	52	25	based	base	VERB
esrj-89750	52	26	on	on	ADP
esrj-89750	52	27	deep	deep	ADJ
esrj-89750	52	28	learning	learning	NOUN
esrj-89750	52	29	include	include	VERB
esrj-89750	52	30	mean	mean	ADJ
esrj-89750	52	31	filtering	filter	VERB
esrj-89750	52	32	,	,	PUNCT
esrj-89750	52	33	median	median	ADJ
esrj-89750	52	34	filtering	filtering	NOUN
esrj-89750	52	35	,	,	PUNCT
esrj-89750	52	36	wavelet	wavelet	NOUN
esrj-89750	52	37	transform	transform	NOUN
esrj-89750	52	38	and	and	CCONJ
esrj-89750	52	39	neighborhood	neighborhood	NOUN
esrj-89750	52	40	averaging	averaging	NOUN
esrj-89750	52	41	.	.	PUNCT
esrj-89750	53	1	wavelet	wavelet	NOUN
esrj-89750	53	2	transform	transform	NOUN
esrj-89750	53	3	is	be	AUX
esrj-89750	53	4	a	a	DET
esrj-89750	53	5	commonly	commonly	ADV
esrj-89750	53	6	used	use	VERB
esrj-89750	53	7	method	method	NOUN
esrj-89750	53	8	at	at	ADP
esrj-89750	53	9	present	present	NOUN
esrj-89750	53	10	(	(	PUNCT
esrj-89750	53	11	li	li	PROPN
esrj-89750	53	12	et	et	PROPN
esrj-89750	53	13	al	al	PROPN
esrj-89750	53	14	.	.	PROPN
esrj-89750	53	15	,	,	PUNCT
esrj-89750	53	16	2017	2017	NUM
esrj-89750	53	17	)	)	PUNCT
esrj-89750	53	18	.	.	PUNCT
esrj-89750	54	1	the	the	DET
esrj-89750	54	2	basic	basic	ADJ
esrj-89750	54	3	principles	principle	NOUN
esrj-89750	54	4	are	be	AUX
esrj-89750	54	5	as	as	SCONJ
esrj-89750	54	6	follows	follow	VERB
esrj-89750	54	7	:	:	PUNCT
esrj-89750	54	8	the	the	DET
esrj-89750	54	9	multi	multi	ADJ
esrj-89750	54	10	-	-	ADJ
esrj-89750	54	11	scale	scale	ADJ
esrj-89750	54	12	wavelet	wavelet	NOUN
esrj-89750	54	13	transform	transform	NOUN
esrj-89750	54	14	is	be	AUX
esrj-89750	54	15	performed	perform	VERB
esrj-89750	54	16	on	on	ADP
esrj-89750	54	17	the	the	DET
esrj-89750	54	18	noisy	noisy	ADJ
esrj-89750	54	19	signal	signal	NOUN
esrj-89750	54	20	,	,	PUNCT
esrj-89750	54	21	and	and	CCONJ
esrj-89750	54	22	the	the	DET
esrj-89750	54	23	wavelet	wavelet	NOUN
esrj-89750	54	24	coefficients	coefficient	NOUN
esrj-89750	54	25	belonging	belong	VERB
esrj-89750	54	26	to	to	ADP
esrj-89750	54	27	the	the	DET
esrj-89750	54	28	noise	noise	NOUN
esrj-89750	54	29	are	be	AUX
esrj-89750	54	30	removed	remove	VERB
esrj-89750	54	31	at	at	ADP
esrj-89750	54	32	each	each	DET
esrj-89750	54	33	scale	scale	NOUN
esrj-89750	54	34	,	,	PUNCT
esrj-89750	54	35	and	and	CCONJ
esrj-89750	54	36	the	the	DET
esrj-89750	54	37	wavelet	wavelet	NOUN
esrj-89750	54	38	coefficients	coefficient	NOUN
esrj-89750	54	39	belonging	belong	VERB
esrj-89750	54	40	to	to	ADP
esrj-89750	54	41	the	the	DET
esrj-89750	54	42	signal	signal	NOUN
esrj-89750	54	43	are	be	AUX
esrj-89750	54	44	preserved	preserve	VERB
esrj-89750	54	45	and	and	CCONJ
esrj-89750	54	46	enhanced	enhance	VERB
esrj-89750	54	47	.	.	PUNCT
esrj-89750	55	1	finally	finally	ADV
esrj-89750	55	2	,	,	PUNCT
esrj-89750	55	3	the	the	DET
esrj-89750	55	4	wavelet	wavelet	NOUN
esrj-89750	55	5	transform	transform	NOUN
esrj-89750	55	6	is	be	AUX
esrj-89750	55	7	used	use	VERB
esrj-89750	55	8	to	to	PART
esrj-89750	55	9	restore	restore	VERB
esrj-89750	55	10	the	the	DET
esrj-89750	55	11	original	original	ADJ
esrj-89750	55	12	signal	signal	NOUN
esrj-89750	55	13	,	,	PUNCT
esrj-89750	55	14	so	so	SCONJ
esrj-89750	55	15	that	that	SCONJ
esrj-89750	55	16	denoising	denoising	NOUN
esrj-89750	55	17	is	be	AUX
esrj-89750	55	18	achieved	achieve	VERB
esrj-89750	55	19	,	,	PUNCT
esrj-89750	55	20	as	as	SCONJ
esrj-89750	55	21	shown	show	VERB
esrj-89750	55	22	in	in	ADP
esrj-89750	55	23	figure	figure	NOUN
esrj-89750	55	24	4	4	NUM
esrj-89750	55	25	.	.	PUNCT
esrj-89750	55	26	figure	figure	VERB
esrj-89750	55	27	4	4	NUM
esrj-89750	55	28	.	.	PUNCT
esrj-89750	55	29	wavelet	wavelet	NOUN
esrj-89750	55	30	denoising	denoising	NOUN
esrj-89750	55	31	(	(	PUNCT
esrj-89750	55	32	3	3	X
esrj-89750	55	33	)	)	PUNCT
esrj-89750	55	34	image	image	NOUN
esrj-89750	55	35	correction	correction	NOUN
esrj-89750	55	36	in	in	ADP
esrj-89750	55	37	the	the	DET
esrj-89750	55	38	acquisition	acquisition	NOUN
esrj-89750	55	39	process	process	NOUN
esrj-89750	55	40	of	of	ADP
esrj-89750	55	41	soil	soil	NOUN
esrj-89750	55	42	remote	remote	ADJ
esrj-89750	55	43	sensing	sense	VERB
esrj-89750	55	44	image	image	NOUN
esrj-89750	55	45	,	,	PUNCT
esrj-89750	55	46	besides	besides	SCONJ
esrj-89750	55	47	the	the	DET
esrj-89750	55	48	image	image	NOUN
esrj-89750	55	49	quality	quality	NOUN
esrj-89750	55	50	degraded	degrade	VERB
esrj-89750	55	51	by	by	ADP
esrj-89750	55	52	noise	noise	NOUN
esrj-89750	55	53	,	,	PUNCT
esrj-89750	55	54	it	it	PRON
esrj-89750	55	55	will	will	AUX
esrj-89750	55	56	also	also	ADV
esrj-89750	55	57	be	be	AUX
esrj-89750	55	58	affected	affect	VERB
esrj-89750	55	59	by	by	ADP
esrj-89750	55	60	the	the	DET
esrj-89750	55	61	characteristics	characteristic	NOUN
esrj-89750	55	62	of	of	ADP
esrj-89750	55	63	the	the	DET
esrj-89750	55	64	sensor	sensor	NOUN
esrj-89750	55	65	itself	itself	PRON
esrj-89750	55	66	,	,	PUNCT
esrj-89750	55	67	the	the	DET
esrj-89750	55	68	illumination	illumination	NOUN
esrj-89750	55	69	conditions	condition	NOUN
esrj-89750	55	70	of	of	ADP
esrj-89750	55	71	the	the	DET
esrj-89750	55	72	ground	ground	NOUN
esrj-89750	55	73	objects	object	NOUN
esrj-89750	55	74	(	(	PUNCT
esrj-89750	55	75	topography	topography	NOUN
esrj-89750	55	76	and	and	CCONJ
esrj-89750	55	77	solar	solar	ADJ
esrj-89750	55	78	altitude	altitude	NOUN
esrj-89750	55	79	angle	angle	NOUN
esrj-89750	55	80	)	)	PUNCT
esrj-89750	55	81	and	and	CCONJ
esrj-89750	55	82	the	the	DET
esrj-89750	55	83	atmospheric	atmospheric	ADJ
esrj-89750	55	84	effect	effect	NOUN
esrj-89750	55	85	,	,	PUNCT
esrj-89750	55	86	which	which	PRON
esrj-89750	55	87	will	will	AUX
esrj-89750	55	88	lead	lead	VERB
esrj-89750	55	89	to	to	ADP
esrj-89750	55	90	the	the	DET
esrj-89750	55	91	inconsistency	inconsistency	NOUN
esrj-89750	55	92	between	between	ADP
esrj-89750	55	93	the	the	DET
esrj-89750	55	94	measured	measured	ADJ
esrj-89750	55	95	values	value	NOUN
esrj-89750	55	96	of	of	ADP
esrj-89750	55	97	remote	remote	ADJ
esrj-89750	55	98	sensing	sense	VERB
esrj-89750	55	99	equipment	equipment	NOUN
esrj-89750	55	100	and	and	CCONJ
esrj-89750	55	101	the	the	DET
esrj-89750	55	102	actual	actual	ADJ
esrj-89750	55	103	spectral	spectral	ADJ
esrj-89750	55	104	emissivity	emissivity	NOUN
esrj-89750	55	105	of	of	ADP
esrj-89750	55	106	the	the	DET
esrj-89750	55	107	ground	ground	NOUN
esrj-89750	55	108	objects	object	NOUN
esrj-89750	55	109	,	,	PUNCT
esrj-89750	55	110	i.e.	i.e.	X
esrj-89750	55	111	,	,	PUNCT
esrj-89750	55	112	radiation	radiation	NOUN
esrj-89750	55	113	distortion	distortion	NOUN
esrj-89750	55	114	.	.	PUNCT
esrj-89750	56	1	the	the	DET
esrj-89750	56	2	radiation	radiation	NOUN
esrj-89750	56	3	distortion	distortion	NOUN
esrj-89750	56	4	will	will	AUX
esrj-89750	56	5	then	then	ADV
esrj-89750	56	6	cause	cause	VERB
esrj-89750	56	7	the	the	DET
esrj-89750	56	8	geometric	geometric	ADJ
esrj-89750	56	9	position	position	NOUN
esrj-89750	56	10	,	,	PUNCT
esrj-89750	56	11	shape	shape	NOUN
esrj-89750	56	12	,	,	PUNCT
esrj-89750	56	13	size	size	NOUN
esrj-89750	56	14	,	,	PUNCT
esrj-89750	56	15	orientation	orientation	NOUN
esrj-89750	56	16	and	and	CCONJ
esrj-89750	56	17	other	other	ADJ
esrj-89750	56	18	features	feature	NOUN
esrj-89750	56	19	of	of	ADP
esrj-89750	56	20	the	the	DET
esrj-89750	56	21	original	original	ADJ
esrj-89750	56	22	image	image	NOUN
esrj-89750	56	23	to	to	PART
esrj-89750	56	24	deviate	deviate	VERB
esrj-89750	56	25	from	from	ADP
esrj-89750	56	26	the	the	DET
esrj-89750	56	27	expression	expression	NOUN
esrj-89750	56	28	requirements	requirement	NOUN
esrj-89750	56	29	in	in	ADP
esrj-89750	56	30	the	the	DET
esrj-89750	56	31	reference	reference	NOUN
esrj-89750	56	32	system	system	NOUN
esrj-89750	56	33	,	,	PUNCT
esrj-89750	56	34	that	that	ADV
esrj-89750	56	35	is	is	ADV
esrj-89750	56	36	,	,	PUNCT
esrj-89750	56	37	geometric	geometric	ADJ
esrj-89750	56	38	distortion	distortion	NOUN
esrj-89750	56	39	(	(	PUNCT
esrj-89750	56	40	ma	ma	PROPN
esrj-89750	56	41	et	et	PROPN
esrj-89750	56	42	al	al	PROPN
esrj-89750	56	43	.	.	PROPN
esrj-89750	56	44	,	,	PUNCT
esrj-89750	56	45	2018	2018	NUM
esrj-89750	56	46	)	)	PUNCT
esrj-89750	56	47	.	.	PUNCT
esrj-89750	57	1	in	in	ADP
esrj-89750	57	2	view	view	NOUN
esrj-89750	57	3	of	of	ADP
esrj-89750	57	4	the	the	DET
esrj-89750	57	5	above	above	ADJ
esrj-89750	57	6	two	two	NUM
esrj-89750	57	7	distortion	distortion	NOUN
esrj-89750	57	8	phenomena	phenomenon	NOUN
esrj-89750	57	9	,	,	PUNCT
esrj-89750	57	10	it	it	PRON
esrj-89750	57	11	is	be	AUX
esrj-89750	57	12	necessary	necessary	ADJ
esrj-89750	57	13	to	to	PART
esrj-89750	57	14	correct	correct	VERB
esrj-89750	57	15	and	and	CCONJ
esrj-89750	57	16	restore	restore	VERB
esrj-89750	57	17	the	the	DET
esrj-89750	57	18	image	image	NOUN
esrj-89750	57	19	.	.	PUNCT
esrj-89750	58	1	1	1	X
esrj-89750	58	2	)	)	PUNCT
esrj-89750	58	3	radiation	radiation	NOUN
esrj-89750	58	4	distortion	distortion	NOUN
esrj-89750	58	5	correction	correction	NOUN
esrj-89750	58	6	.	.	PUNCT
esrj-89750	59	1	according	accord	VERB
esrj-89750	59	2	to	to	ADP
esrj-89750	59	3	the	the	DET
esrj-89750	59	4	three	three	NUM
esrj-89750	59	5	causes	cause	NOUN
esrj-89750	59	6	of	of	ADP
esrj-89750	59	7	radiation	radiation	NOUN
esrj-89750	59	8	distortion	distortion	NOUN
esrj-89750	59	9	,	,	PUNCT
esrj-89750	59	10	i.e.	i.e.	X
esrj-89750	59	11	the	the	DET
esrj-89750	59	12	characteristics	characteristic	NOUN
esrj-89750	59	13	of	of	ADP
esrj-89750	59	14	the	the	DET
esrj-89750	59	15	sensor	sensor	NOUN
esrj-89750	59	16	itself	itself	PRON
esrj-89750	59	17	,	,	PUNCT
esrj-89750	59	18	the	the	DET
esrj-89750	59	19	illumination	illumination	NOUN
esrj-89750	59	20	conditions	condition	NOUN
esrj-89750	59	21	of	of	ADP
esrj-89750	59	22	the	the	DET
esrj-89750	59	23	ground	ground	NOUN
esrj-89750	59	24	objects	object	NOUN
esrj-89750	59	25	and	and	CCONJ
esrj-89750	59	26	the	the	DET
esrj-89750	59	27	atmospheric	atmospheric	ADJ
esrj-89750	59	28	effect	effect	NOUN
esrj-89750	59	29	.	.	PUNCT
esrj-89750	60	1	the	the	DET
esrj-89750	60	2	calibration	calibration	NOUN
esrj-89750	60	3	of	of	ADP
esrj-89750	60	4	the	the	DET
esrj-89750	60	5	remote	remote	ADJ
esrj-89750	60	6	sensor	sensor	NOUN
esrj-89750	60	7	,	,	PUNCT
esrj-89750	60	8	the	the	DET
esrj-89750	60	9	solar	solar	ADJ
esrj-89750	60	10	altitude	altitude	NOUN
esrj-89750	60	11	and	and	CCONJ
esrj-89750	60	12	the	the	DET
esrj-89750	60	13	terrain	terrain	NOUN
esrj-89750	60	14	,	,	PUNCT
esrj-89750	60	15	and	and	CCONJ
esrj-89750	60	16	the	the	DET
esrj-89750	60	17	atmospheric	atmospheric	ADJ
esrj-89750	60	18	correction	correction	NOUN
esrj-89750	60	19	are	be	AUX
esrj-89750	60	20	carried	carry	VERB
esrj-89750	60	21	out	out	ADP
esrj-89750	60	22	by	by	ADP
esrj-89750	60	23	these	these	DET
esrj-89750	60	24	three	three	NUM
esrj-89750	60	25	methods	method	NOUN
esrj-89750	60	26	,	,	PUNCT
esrj-89750	60	27	respectively	respectively	ADV
esrj-89750	60	28	,	,	PUNCT
esrj-89750	60	29	as	as	SCONJ
esrj-89750	60	30	shown	show	VERB
esrj-89750	60	31	in	in	ADP
esrj-89750	60	32	preprocessing	preprocessing	NOUN
esrj-89750	60	33	of	of	ADP
esrj-89750	60	34	soil	soil	NOUN
esrj-89750	60	35	remote	remote	ADJ
esrj-89750	60	36	sensing	sensing	NOUN
esrj-89750	60	37	images	image	NOUN
esrj-89750	60	38	as	as	SCONJ
esrj-89750	60	39	shown	show	VERB
esrj-89750	60	40	in	in	ADP
esrj-89750	60	41	figure	figure	NOUN
esrj-89750	60	42	2	2	NUM
esrj-89750	60	43	,	,	PUNCT
esrj-89750	60	44	the	the	DET
esrj-89750	60	45	quality	quality	NOUN
esrj-89750	60	46	of	of	ADP
esrj-89750	60	47	the	the	DET
esrj-89750	60	48	original	original	ADJ
esrj-89750	60	49	soil	soil	NOUN
esrj-89750	60	50	remote	remote	ADJ
esrj-89750	60	51	sensing	sense	VERB
esrj-89750	60	52	image	image	NOUN
esrj-89750	60	53	is	be	AUX
esrj-89750	60	54	not	not	PART
esrj-89750	60	55	high	high	ADJ
esrj-89750	60	56	,	,	PUNCT
esrj-89750	60	57	which	which	PRON
esrj-89750	60	58	is	be	AUX
esrj-89750	60	59	not	not	PART
esrj-89750	60	60	conducive	conducive	ADJ
esrj-89750	60	61	to	to	ADP
esrj-89750	60	62	the	the	DET
esrj-89750	60	63	subsequent	subsequent	ADJ
esrj-89750	60	64	image	image	NOUN
esrj-89750	60	65	classification	classification	NOUN
esrj-89750	60	66	.	.	PUNCT
esrj-89750	61	1	therefore	therefore	ADV
esrj-89750	61	2	,	,	PUNCT
esrj-89750	61	3	it	it	PRON
esrj-89750	61	4	is	be	AUX
esrj-89750	61	5	necessary	necessary	ADJ
esrj-89750	61	6	to	to	AUX
esrj-89750	61	7	pre	pre	VERB
esrj-89750	61	8	-	-	VERB
esrj-89750	61	9	process	process	VERB
esrj-89750	61	10	the	the	DET
esrj-89750	61	11	image	image	NOUN
esrj-89750	61	12	before	before	SCONJ
esrj-89750	61	13	classification	classification	NOUN
esrj-89750	61	14	to	to	PART
esrj-89750	61	15	improve	improve	VERB
esrj-89750	61	16	the	the	DET
esrj-89750	61	17	image	image	NOUN
esrj-89750	61	18	quality	quality	NOUN
esrj-89750	61	19	,	,	PUNCT
esrj-89750	61	20	including	include	VERB
esrj-89750	61	21	image	image	NOUN
esrj-89750	61	22	graying	graying	NOUN
esrj-89750	61	23	,	,	PUNCT
esrj-89750	61	24	image	image	NOUN
esrj-89750	61	25	denoising	denoising	NOUN
esrj-89750	61	26	and	and	CCONJ
esrj-89750	61	27	image	image	NOUN
esrj-89750	61	28	correction	correction	NOUN
esrj-89750	61	29	(	(	PUNCT
esrj-89750	61	30	he	he	PRON
esrj-89750	61	31	et	et	PROPN
esrj-89750	61	32	al	al	PROPN
esrj-89750	61	33	.	.	PROPN
esrj-89750	61	34	,	,	PUNCT
esrj-89750	61	35	2018	2018	NUM
esrj-89750	61	36	)	)	PUNCT
esrj-89750	61	37	.	.	PUNCT
esrj-89750	62	1	(	(	PUNCT
esrj-89750	62	2	1	1	X
esrj-89750	62	3	)	)	PUNCT
esrj-89750	62	4	image	image	NOUN
esrj-89750	62	5	graying	gray	VERB
esrj-89750	62	6	grayscale	grayscale	NOUN
esrj-89750	62	7	image	image	NOUN
esrj-89750	62	8	,	,	PUNCT
esrj-89750	62	9	commonly	commonly	ADV
esrj-89750	62	10	speaking	speak	VERB
esrj-89750	62	11	,	,	PUNCT
esrj-89750	62	12	is	be	AUX
esrj-89750	62	13	the	the	DET
esrj-89750	62	14	conversion	conversion	NOUN
esrj-89750	62	15	of	of	ADP
esrj-89750	62	16	a	a	DET
esrj-89750	62	17	color	color	NOUN
esrj-89750	62	18	image	image	NOUN
esrj-89750	62	19	to	to	ADP
esrj-89750	62	20	a	a	DET
esrj-89750	62	21	gray	gray	ADJ
esrj-89750	62	22	image	image	NOUN
esrj-89750	62	23	whose	whose	DET
esrj-89750	62	24	pixel	pixel	NOUN
esrj-89750	62	25	value	value	NOUN
esrj-89750	62	26	is	be	AUX
esrj-89750	62	27	between	between	ADP
esrj-89750	62	28	0	0	NUM
esrj-89750	62	29	and	and	CCONJ
esrj-89750	62	30	255	255	NUM
esrj-89750	62	31	.	.	PUNCT
esrj-89750	63	1	the	the	DET
esrj-89750	63	2	whole	whole	ADJ
esrj-89750	63	3	image	image	NOUN
esrj-89750	63	4	is	be	AUX
esrj-89750	63	5	composed	compose	VERB
esrj-89750	63	6	of	of	ADP
esrj-89750	63	7	different	different	ADJ
esrj-89750	63	8	degrees	degree	NOUN
esrj-89750	63	9	of	of	ADP
esrj-89750	63	10	gray	gray	ADJ
esrj-89750	63	11	.	.	PUNCT
esrj-89750	64	1	its	its	PRON
esrj-89750	64	2	purpose	purpose	NOUN
esrj-89750	64	3	is	be	AUX
esrj-89750	64	4	to	to	PART
esrj-89750	64	5	reduce	reduce	VERB
esrj-89750	64	6	the	the	DET
esrj-89750	64	7	interference	interference	NOUN
esrj-89750	64	8	of	of	ADP
esrj-89750	64	9	color	color	NOUN
esrj-89750	64	10	to	to	ADP
esrj-89750	64	11	the	the	DET
esrj-89750	64	12	target	target	NOUN
esrj-89750	64	13	information	information	NOUN
esrj-89750	64	14	in	in	ADP
esrj-89750	64	15	the	the	DET
esrj-89750	64	16	image	image	NOUN
esrj-89750	64	17	.	.	PUNCT
esrj-89750	65	1	there	there	PRON
esrj-89750	65	2	are	be	VERB
esrj-89750	65	3	four	four	NUM
esrj-89750	65	4	main	main	ADJ
esrj-89750	65	5	methods	method	NOUN
esrj-89750	65	6	of	of	ADP
esrj-89750	65	7	gray	gray	ADJ
esrj-89750	65	8	image	image	NOUN
esrj-89750	65	9	processing	processing	NOUN
esrj-89750	65	10	:	:	PUNCT
esrj-89750	65	11	component	component	NOUN
esrj-89750	65	12	method	method	NOUN
esrj-89750	65	13	,	,	PUNCT
esrj-89750	65	14	maximum	maximum	ADJ
esrj-89750	65	15	method	method	NOUN
esrj-89750	65	16	,	,	PUNCT
esrj-89750	65	17	average	average	ADJ
esrj-89750	65	18	method	method	NOUN
esrj-89750	65	19	,	,	PUNCT
esrj-89750	65	20	and	and	CCONJ
esrj-89750	65	21	weighted	weight	VERB
esrj-89750	65	22	average	average	ADJ
esrj-89750	65	23	method	method	NOUN
esrj-89750	65	24	,	,	PUNCT
esrj-89750	65	25	as	as	SCONJ
esrj-89750	65	26	shown	show	VERB
esrj-89750	65	27	in	in	ADP
esrj-89750	65	28	table	table	NOUN
esrj-89750	65	29	1	1	NUM
esrj-89750	65	30	.	.	PUNCT
esrj-89750	66	1	(	(	PUNCT
esrj-89750	66	2	2	2	X
esrj-89750	66	3	)	)	PUNCT
esrj-89750	66	4	image	image	NOUN
esrj-89750	66	5	denoising	denoising	NOUN
esrj-89750	66	6	influenced	influence	VERB
esrj-89750	66	7	by	by	ADP
esrj-89750	66	8	natural	natural	ADJ
esrj-89750	66	9	factors	factor	NOUN
esrj-89750	66	10	such	such	ADJ
esrj-89750	66	11	as	as	ADP
esrj-89750	66	12	illumination	illumination	NOUN
esrj-89750	66	13	,	,	PUNCT
esrj-89750	66	14	cloud	cloud	NOUN
esrj-89750	66	15	,	,	PUNCT
esrj-89750	66	16	and	and	CCONJ
esrj-89750	66	17	equipment	equipment	NOUN
esrj-89750	66	18	itself	itself	PRON
esrj-89750	66	19	,	,	PUNCT
esrj-89750	66	20	the	the	DET
esrj-89750	66	21	original	original	ADJ
esrj-89750	66	22	remote	remote	ADJ
esrj-89750	66	23	sensing	sense	VERB
esrj-89750	66	24	image	image	NOUN
esrj-89750	66	25	collected	collect	VERB
esrj-89750	66	26	contains	contain	VERB
esrj-89750	66	27	a	a	DET
esrj-89750	66	28	lot	lot	NOUN
esrj-89750	66	29	of	of	ADP
esrj-89750	66	30	noise	noise	NOUN
esrj-89750	66	31	.	.	PUNCT
esrj-89750	67	1	after	after	ADP
esrj-89750	67	2	graying	gray	VERB
esrj-89750	67	3	,	,	PUNCT
esrj-89750	67	4	it	it	PRON
esrj-89750	67	5	can	can	AUX
esrj-89750	67	6	see	see	VERB
esrj-89750	67	7	that	that	SCONJ
esrj-89750	67	8	there	there	PRON
esrj-89750	67	9	are	be	VERB
esrj-89750	67	10	many	many	ADJ
esrj-89750	67	11	white	white	ADJ
esrj-89750	67	12	elements	element	NOUN
esrj-89750	67	13	on	on	ADP
esrj-89750	67	14	the	the	DET
esrj-89750	67	15	image	image	NOUN
esrj-89750	67	16	.	.	PUNCT
esrj-89750	68	1	these	these	DET
esrj-89750	68	2	elements	element	NOUN
esrj-89750	68	3	are	be	AUX
esrj-89750	68	4	image	image	NOUN
esrj-89750	68	5	noise	noise	NOUN
esrj-89750	68	6	,	,	PUNCT
esrj-89750	68	7	as	as	SCONJ
esrj-89750	68	8	shown	show	VERB
esrj-89750	68	9	in	in	ADP
esrj-89750	68	10	figure	figure	NOUN
esrj-89750	68	11	3	3	NUM
esrj-89750	68	12	.	.	PUNCT
esrj-89750	68	13	figure	figure	NOUN
esrj-89750	68	14	3	3	NUM
esrj-89750	68	15	.	.	PUNCT
esrj-89750	68	16	image	image	NOUN
esrj-89750	68	17	noise	noise	VERB
esrj-89750	68	18	the	the	DET
esrj-89750	68	19	existence	existence	NOUN
esrj-89750	68	20	of	of	ADP
esrj-89750	68	21	image	image	NOUN
esrj-89750	68	22	noise	noise	NOUN
esrj-89750	68	23	will	will	AUX
esrj-89750	68	24	blur	blur	VERB
esrj-89750	68	25	the	the	DET
esrj-89750	68	26	image	image	NOUN
esrj-89750	68	27	target	target	NOUN
esrj-89750	68	28	and	and	CCONJ
esrj-89750	68	29	reduce	reduce	VERB
esrj-89750	68	30	the	the	DET
esrj-89750	68	31	image	image	NOUN
esrj-89750	68	32	quality	quality	NOUN
esrj-89750	68	33	,	,	PUNCT
esrj-89750	68	34	so	so	SCONJ
esrj-89750	68	35	it	it	PRON
esrj-89750	68	36	needs	need	VERB
esrj-89750	68	37	to	to	PART
esrj-89750	68	38	be	be	AUX
esrj-89750	68	39	de	de	ADJ
esrj-89750	68	40	-	-	ADJ
esrj-89750	68	41	noised	noised	ADJ
esrj-89750	68	42	.	.	PUNCT
esrj-89750	69	1	at	at	ADP
esrj-89750	69	2	present	present	ADJ
esrj-89750	69	3	,	,	PUNCT
esrj-89750	69	4	denoising	denoising	NOUN
esrj-89750	69	5	methods	method	NOUN
esrj-89750	69	6	table	table	NOUN
esrj-89750	69	7	1	1	NUM
esrj-89750	69	8	.	.	X
esrj-89750	70	1	four	four	NUM
esrj-89750	70	2	methods	method	NOUN
esrj-89750	70	3	of	of	ADP
esrj-89750	70	4	image	image	NOUN
esrj-89750	70	5	graying	graying	NOUN
esrj-89750	70	6	method	method	NOUN
esrj-89750	70	7	definition	definition	NOUN
esrj-89750	70	8	formula	formula	NOUN
esrj-89750	70	9	explanation	explanation	NOUN
esrj-89750	70	10	of	of	ADP
esrj-89750	70	11	parameters	parameter	NOUN
esrj-89750	70	12	component	component	NOUN
esrj-89750	70	13	method	method	NOUN
esrj-89750	70	14	taking	take	VERB
esrj-89750	70	15	the	the	DET
esrj-89750	70	16	brightness	brightness	NOUN
esrj-89750	70	17	of	of	ADP
esrj-89750	70	18	three	three	NUM
esrj-89750	70	19	components	component	NOUN
esrj-89750	70	20	in	in	ADP
esrj-89750	70	21	a	a	DET
esrj-89750	70	22	color	color	NOUN
esrj-89750	70	23	image	image	NOUN
esrj-89750	70	24	as	as	ADP
esrj-89750	70	25	the	the	DET
esrj-89750	70	26	gray	gray	ADJ
esrj-89750	70	27	value	value	NOUN
esrj-89750	70	28	of	of	ADP
esrj-89750	70	29	three	three	NUM
esrj-89750	70	30	gray	gray	ADJ
esrj-89750	70	31	images	image	NOUN
esrj-89750	70	32	,	,	PUNCT
esrj-89750	70	33	a	a	DET
esrj-89750	70	34	gray	gray	ADJ
esrj-89750	70	35	image	image	NOUN
esrj-89750	70	36	can	can	AUX
esrj-89750	70	37	be	be	AUX
esrj-89750	70	38	selected	select	VERB
esrj-89750	70	39	according	accord	VERB
esrj-89750	70	40	to	to	ADP
esrj-89750	70	41	the	the	DET
esrj-89750	70	42	application	application	NOUN
esrj-89750	70	43	needs	need	NOUN
esrj-89750	70	44	.	.	PUNCT
esrj-89750	71	1	f	f	X
esrj-89750	72	1	i	i	PRON
esrj-89750	72	2	j	j	NOUN
esrj-89750	73	1	r	r	NOUN
esrj-89750	73	2	i	i	PRON
esrj-89750	73	3	j1	j1	PROPN
esrj-89750	73	4	,	,	PUNCT
esrj-89750	73	5	,	,	PUNCT
esrj-89750	73	6	(	(	PUNCT
esrj-89750	73	7	)	)	PUNCT
esrj-89750	73	8	=	=	SYM
esrj-89750	74	1	(	(	PUNCT
esrj-89750	74	2	)	)	PUNCT
esrj-89750	74	3	f	f	PROPN
esrj-89750	75	1	i	i	PRON
esrj-89750	75	2	j	j	PROPN
esrj-89750	76	1	g	g	PROPN
esrj-89750	76	2	i	i	PROPN
esrj-89750	76	3	j2	j2	PROPN
esrj-89750	76	4	,	,	PUNCT
esrj-89750	76	5	,	,	PUNCT
esrj-89750	76	6	(	(	PUNCT
esrj-89750	76	7	)	)	PUNCT
esrj-89750	76	8	=	=	SYM
esrj-89750	76	9	(	(	PUNCT
esrj-89750	76	10	)	)	PUNCT
esrj-89750	76	11	f	f	PROPN
esrj-89750	77	1	i	i	PRON
esrj-89750	77	2	j	j	PROPN
esrj-89750	78	1	b	b	VERB
esrj-89750	78	2	i	i	PROPN
esrj-89750	78	3	j3	j3	PROPN
esrj-89750	78	4	,	,	PUNCT
esrj-89750	78	5	,	,	PUNCT
esrj-89750	78	6	(	(	PUNCT
esrj-89750	78	7	)	)	PUNCT
esrj-89750	78	8	=	=	SYM
esrj-89750	78	9	(	(	PUNCT
esrj-89750	78	10	)	)	PUNCT
esrj-89750	78	11	(	(	PUNCT
esrj-89750	78	12	1	1	X
esrj-89750	78	13	)	)	PUNCT
esrj-89750	78	14	f	f	NOUN
esrj-89750	79	1	i	i	PRON
esrj-89750	79	2	j	j	PROPN
esrj-89750	79	3	kk	kk	INTJ
esrj-89750	79	4	,	,	PUNCT
esrj-89750	79	5	,	,	PUNCT
esrj-89750	79	6	,	,	PUNCT
esrj-89750	79	7	(	(	PUNCT
esrj-89750	79	8	)	)	PUNCT
esrj-89750	79	9	=(	=(	NOUN
esrj-89750	79	10	)	)	PUNCT
esrj-89750	79	11	1	1	NUM
esrj-89750	79	12	2	2	NUM
esrj-89750	79	13	3	3	NUM
esrj-89750	79	14	is	be	AUX
esrj-89750	79	15	the	the	DET
esrj-89750	79	16	gray	gray	ADJ
esrj-89750	79	17	value	value	NOUN
esrj-89750	79	18	of	of	ADP
esrj-89750	79	19	the	the	DET
esrj-89750	79	20	converted	convert	VERB
esrj-89750	79	21	gray	gray	ADJ
esrj-89750	79	22	image	image	NOUN
esrj-89750	79	23	at	at	ADP
esrj-89750	79	24	(	(	PUNCT
esrj-89750	79	25	i	i	PROPN
esrj-89750	79	26	,	,	PUNCT
esrj-89750	79	27	j	j	PROPN
esrj-89750	79	28	)	)	PUNCT
esrj-89750	79	29	;	;	PUNCT
esrj-89750	79	30	r	r	X
esrj-89750	79	31	,	,	PUNCT
esrj-89750	79	32	g	g	NOUN
esrj-89750	79	33	,	,	PUNCT
esrj-89750	79	34	and	and	CCONJ
esrj-89750	79	35	b	b	X
esrj-89750	79	36	represent	represent	VERB
esrj-89750	79	37	color	color	NOUN
esrj-89750	79	38	components	component	NOUN
esrj-89750	79	39	,	,	PUNCT
esrj-89750	79	40	ranging	range	VERB
esrj-89750	79	41	from	from	ADP
esrj-89750	79	42	0	0	NUM
esrj-89750	79	43	to	to	ADP
esrj-89750	79	44	255	255	NUM
esrj-89750	79	45	maximum	maximum	ADJ
esrj-89750	79	46	method	method	NOUN
esrj-89750	79	47	the	the	DET
esrj-89750	79	48	maximum	maximum	ADJ
esrj-89750	79	49	brightness	brightness	NOUN
esrj-89750	79	50	of	of	ADP
esrj-89750	79	51	three	three	NUM
esrj-89750	79	52	components	component	NOUN
esrj-89750	79	53	in	in	ADP
esrj-89750	79	54	a	a	DET
esrj-89750	79	55	color	color	NOUN
esrj-89750	79	56	image	image	NOUN
esrj-89750	79	57	is	be	AUX
esrj-89750	79	58	taken	take	VERB
esrj-89750	79	59	as	as	ADP
esrj-89750	79	60	the	the	DET
esrj-89750	79	61	gray	gray	ADJ
esrj-89750	79	62	value	value	NOUN
esrj-89750	79	63	of	of	ADP
esrj-89750	79	64	a	a	DET
esrj-89750	79	65	gray	gray	ADJ
esrj-89750	79	66	image	image	NOUN
esrj-89750	79	67	.	.	PUNCT
esrj-89750	80	1	f	f	X
esrj-89750	81	1	i	i	PRON
esrj-89750	81	2	j	j	NOUN
esrj-89750	82	1	r	r	NOUN
esrj-89750	82	2	i	i	PRON
esrj-89750	83	1	j	j	PROPN
esrj-89750	84	1	g	g	NOUN
esrj-89750	84	2	i	i	PRON
esrj-89750	85	1	j	j	PROPN
esrj-89750	86	1	b	b	PROPN
esrj-89750	86	2	i	i	PRON
esrj-89750	86	3	j	j	PROPN
esrj-89750	86	4	,	,	PUNCT
esrj-89750	86	5	max	max	PROPN
esrj-89750	86	6	,	,	PUNCT
esrj-89750	86	7	,	,	PUNCT
esrj-89750	86	8	,	,	PUNCT
esrj-89750	86	9	,	,	PUNCT
esrj-89750	86	10	,	,	PUNCT
esrj-89750	86	11	(	(	PUNCT
esrj-89750	86	12	)	)	PUNCT
esrj-89750	86	13	=	=	SYM
esrj-89750	86	14	(	(	PUNCT
esrj-89750	86	15	)	)	PUNCT
esrj-89750	86	16	(	(	PUNCT
esrj-89750	86	17	)	)	PUNCT
esrj-89750	86	18	(	(	PUNCT
esrj-89750	86	19	)	)	PUNCT
esrj-89750	86	20			NOUN
esrj-89750	86	21			NOUN
esrj-89750	86	22	(	(	PUNCT
esrj-89750	86	23	2	2	X
esrj-89750	86	24	)	)	PUNCT
esrj-89750	86	25	average	average	ADJ
esrj-89750	86	26	method	method	NOUN
esrj-89750	86	27	the	the	DET
esrj-89750	86	28	three	three	NUM
esrj-89750	86	29	-	-	PUNCT
esrj-89750	86	30	component	component	NOUN
esrj-89750	86	31	brightness	brightness	NOUN
esrj-89750	86	32	of	of	ADP
esrj-89750	86	33	the	the	DET
esrj-89750	86	34	color	color	NOUN
esrj-89750	86	35	image	image	NOUN
esrj-89750	86	36	is	be	AUX
esrj-89750	86	37	averaged	average	VERB
esrj-89750	86	38	to	to	PART
esrj-89750	86	39	get	get	VERB
esrj-89750	86	40	a	a	DET
esrj-89750	86	41	gray	gray	ADJ
esrj-89750	86	42	value	value	NOUN
esrj-89750	86	43	.	.	PUNCT
esrj-89750	87	1	f	f	X
esrj-89750	88	1	i	i	PRON
esrj-89750	88	2	j	j	NOUN
esrj-89750	89	1	r	r	NOUN
esrj-89750	89	2	i	i	PRON
esrj-89750	90	1	j	j	PROPN
esrj-89750	91	1	g	g	NOUN
esrj-89750	91	2	i	i	PRON
esrj-89750	91	3	j	j	PROPN
esrj-89750	92	1	b	b	NOUN
esrj-89750	92	2	i	i	PRON
esrj-89750	92	3	j	j	PROPN
esrj-89750	92	4	,	,	PUNCT
esrj-89750	92	5	,	,	PUNCT
esrj-89750	92	6	,	,	PUNCT
esrj-89750	92	7	,	,	PUNCT
esrj-89750	92	8	(	(	PUNCT
esrj-89750	92	9	)	)	PUNCT
esrj-89750	92	10	=	=	SYM
esrj-89750	92	11	(	(	PUNCT
esrj-89750	92	12	)	)	PUNCT
esrj-89750	93	1	+	+	CCONJ
esrj-89750	93	2	(	(	PUNCT
esrj-89750	93	3	)	)	PUNCT
esrj-89750	93	4	+	+	CCONJ
esrj-89750	93	5	(	(	PUNCT
esrj-89750	93	6	)	)	PUNCT
esrj-89750	93	7	3	3	NUM
esrj-89750	93	8	(	(	PUNCT
esrj-89750	93	9	3	3	NUM
esrj-89750	93	10	)	)	PUNCT
esrj-89750	93	11	weighted	weight	VERB
esrj-89750	93	12	average	average	ADJ
esrj-89750	93	13	method	method	NOUN
esrj-89750	93	14	.	.	PUNCT
esrj-89750	94	1	according	accord	VERB
esrj-89750	94	2	to	to	ADP
esrj-89750	94	3	the	the	DET
esrj-89750	94	4	importance	importance	NOUN
esrj-89750	94	5	and	and	CCONJ
esrj-89750	94	6	other	other	ADJ
esrj-89750	94	7	indicators	indicator	NOUN
esrj-89750	94	8	,	,	PUNCT
esrj-89750	94	9	the	the	DET
esrj-89750	94	10	three	three	NUM
esrj-89750	94	11	components	component	NOUN
esrj-89750	94	12	are	be	AUX
esrj-89750	94	13	weighted	weight	VERB
esrj-89750	94	14	averaged	average	VERB
esrj-89750	94	15	with	with	ADP
esrj-89750	94	16	different	different	ADJ
esrj-89750	94	17	weights	weight	NOUN
esrj-89750	94	18	.	.	PUNCT
esrj-89750	95	1	f	f	X
esrj-89750	96	1	i	i	PRON
esrj-89750	96	2	j	j	NOUN
esrj-89750	97	1	r	r	NOUN
esrj-89750	97	2	i	i	PRON
esrj-89750	98	1	j	j	PROPN
esrj-89750	99	1	g	g	NOUN
esrj-89750	99	2	i	i	PRON
esrj-89750	99	3	j	j	PROPN
esrj-89750	100	1	b	b	PROPN
esrj-89750	100	2	i	i	PRON
esrj-89750	100	3	j	j	PROPN
esrj-89750	100	4	,	,	PUNCT
esrj-89750	100	5	.	.	PUNCT
esrj-89750	101	1	,	,	PUNCT
esrj-89750	101	2	.	.	PUNCT
esrj-89750	102	1	,	,	PUNCT
esrj-89750	102	2	.	.	PUNCT
esrj-89750	103	1	,	,	PUNCT
esrj-89750	103	2	(	(	PUNCT
esrj-89750	103	3	)	)	PUNCT
esrj-89750	103	4	=	=	SYM
esrj-89750	103	5	(	(	PUNCT
esrj-89750	103	6	)	)	PUNCT
esrj-89750	103	7	+	+	CCONJ
esrj-89750	103	8	(	(	PUNCT
esrj-89750	103	9	)	)	PUNCT
esrj-89750	103	10	+	+	CCONJ
esrj-89750	103	11	(	(	PUNCT
esrj-89750	103	12	)	)	PUNCT
esrj-89750	103	13	0	0	NUM
esrj-89750	104	1	30	30	NUM
esrj-89750	104	2	0	0	NUM
esrj-89750	104	3	59	59	NUM
esrj-89750	104	4	0	0	NUM
esrj-89750	104	5	11	11	NUM
esrj-89750	104	6	(	(	PUNCT
esrj-89750	104	7	4	4	NUM
esrj-89750	104	8	)	)	PUNCT
esrj-89750	104	9	360	360	NUM
esrj-89750	104	10	shujun	shujun	PROPN
esrj-89750	104	11	liang1	liang1	PROPN
esrj-89750	104	12	,	,	PUNCT
esrj-89750	104	13	jing	je	VERB
esrj-89750	105	1	cheng2	cheng2	PROPN
esrj-89750	105	2	*	*	PROPN
esrj-89750	105	3	,	,	PUNCT
esrj-89750	105	4	jianwei	jianwei	PROPN
esrj-89750	105	5	zhang1	zhang1	PROPN
esrj-89750	105	6	figure	figure	NOUN
esrj-89750	105	7	6	6	NUM
esrj-89750	105	8	.	.	PUNCT
esrj-89750	106	1	sensor	sensor	NOUN
esrj-89750	106	2	calibration	calibration	NOUN
esrj-89750	106	3	:	:	PUNCT
esrj-89750	106	4	in	in	ADP
esrj-89750	106	5	order	order	NOUN
esrj-89750	106	6	to	to	PART
esrj-89750	106	7	eliminate	eliminate	VERB
esrj-89750	106	8	the	the	DET
esrj-89750	106	9	radiation	radiation	NOUN
esrj-89750	106	10	error	error	NOUN
esrj-89750	106	11	caused	cause	VERB
esrj-89750	106	12	by	by	ADP
esrj-89750	106	13	the	the	DET
esrj-89750	106	14	sensor	sensor	NOUN
esrj-89750	106	15	itself	itself	PRON
esrj-89750	106	16	,	,	PUNCT
esrj-89750	106	17	the	the	DET
esrj-89750	106	18	dimensionless	dimensionless	NOUN
esrj-89750	106	19	dn	dn	NOUN
esrj-89750	106	20	value	value	NOUN
esrj-89750	106	21	recorded	record	VERB
esrj-89750	106	22	by	by	ADP
esrj-89750	106	23	the	the	DET
esrj-89750	106	24	sensor	sensor	NOUN
esrj-89750	106	25	is	be	AUX
esrj-89750	106	26	converted	convert	VERB
esrj-89750	106	27	into	into	ADP
esrj-89750	106	28	the	the	DET
esrj-89750	106	29	atmospheric	atmospheric	ADJ
esrj-89750	106	30	top	top	ADJ
esrj-89750	106	31	radiation	radiation	NOUN
esrj-89750	106	32	brightness	brightness	NOUN
esrj-89750	106	33	or	or	CCONJ
esrj-89750	106	34	reflectance	reflectance	NOUN
esrj-89750	106	35	with	with	ADP
esrj-89750	106	36	practical	practical	ADJ
esrj-89750	106	37	physical	physical	ADJ
esrj-89750	106	38	significance	significance	NOUN
esrj-89750	106	39	.	.	PUNCT
esrj-89750	107	1	topographic	topographic	ADJ
esrj-89750	107	2	correction	correction	NOUN
esrj-89750	107	3	:	:	PUNCT
esrj-89750	107	4	it	it	PRON
esrj-89750	107	5	can	can	AUX
esrj-89750	107	6	mainly	mainly	ADV
esrj-89750	107	7	reduce	reduce	VERB
esrj-89750	107	8	the	the	DET
esrj-89750	107	9	shadows	shadow	NOUN
esrj-89750	107	10	in	in	ADP
esrj-89750	107	11	remote	remote	ADJ
esrj-89750	107	12	sensing	sensing	NOUN
esrj-89750	107	13	images	image	NOUN
esrj-89750	107	14	,	,	PUNCT
esrj-89750	107	15	because	because	SCONJ
esrj-89750	107	16	it	it	PRON
esrj-89750	107	17	restores	restore	VERB
esrj-89750	107	18	spectral	spectral	ADJ
esrj-89750	107	19	information	information	NOUN
esrj-89750	107	20	,	,	PUNCT
esrj-89750	107	21	but	but	CCONJ
esrj-89750	107	22	because	because	SCONJ
esrj-89750	107	23	the	the	DET
esrj-89750	107	24	existence	existence	NOUN
esrj-89750	107	25	of	of	ADP
esrj-89750	107	26	shadows	shadow	NOUN
esrj-89750	107	27	will	will	AUX
esrj-89750	107	28	make	make	VERB
esrj-89750	107	29	the	the	DET
esrj-89750	107	30	image	image	NOUN
esrj-89750	107	31	stereoscopic	stereoscopic	NOUN
esrj-89750	107	32	,	,	PUNCT
esrj-89750	107	33	which	which	PRON
esrj-89750	107	34	is	be	AUX
esrj-89750	107	35	a	a	DET
esrj-89750	107	36	visual	visual	ADJ
esrj-89750	107	37	experience	experience	NOUN
esrj-89750	107	38	.	.	PUNCT
esrj-89750	108	1	to	to	PART
esrj-89750	108	2	see	see	VERB
esrj-89750	108	3	whether	whether	SCONJ
esrj-89750	108	4	topographic	topographic	ADJ
esrj-89750	108	5	correction	correction	NOUN
esrj-89750	108	6	improves	improve	VERB
esrj-89750	108	7	image	image	NOUN
esrj-89750	108	8	quality	quality	NOUN
esrj-89750	108	9	,	,	PUNCT
esrj-89750	108	10	it	it	PRON
esrj-89750	108	11	depends	depend	VERB
esrj-89750	108	12	on	on	ADP
esrj-89750	108	13	whether	whether	SCONJ
esrj-89750	108	14	spectral	spectral	ADJ
esrj-89750	108	15	information	information	NOUN
esrj-89750	108	16	has	have	AUX
esrj-89750	108	17	been	be	AUX
esrj-89750	108	18	corrected	correct	VERB
esrj-89750	108	19	.	.	PUNCT
esrj-89750	109	1	cosine	cosine	NOUN
esrj-89750	109	2	correction	correction	NOUN
esrj-89750	109	3	and	and	CCONJ
esrj-89750	109	4	semi	semi	ADJ
esrj-89750	109	5	-	-	ADJ
esrj-89750	109	6	empirical	empirical	ADJ
esrj-89750	109	7	c	c	NOUN
esrj-89750	109	8	correction	correction	NOUN
esrj-89750	109	9	are	be	AUX
esrj-89750	109	10	usually	usually	ADV
esrj-89750	109	11	used	use	VERB
esrj-89750	109	12	for	for	ADP
esrj-89750	109	13	topographic	topographic	ADJ
esrj-89750	109	14	correction	correction	NOUN
esrj-89750	109	15	(	(	PUNCT
esrj-89750	109	16	mccord	mccord	PROPN
esrj-89750	109	17	et	et	PROPN
esrj-89750	109	18	al	al	PROPN
esrj-89750	109	19	.	.	PROPN
esrj-89750	109	20	,	,	PUNCT
esrj-89750	109	21	2017	2017	NUM
esrj-89750	109	22	)	)	PUNCT
esrj-89750	109	23	.	.	PUNCT
esrj-89750	110	1	atmospheric	atmospheric	ADJ
esrj-89750	110	2	correction	correction	NOUN
esrj-89750	110	3	is	be	AUX
esrj-89750	110	4	divided	divide	VERB
esrj-89750	110	5	into	into	ADP
esrj-89750	110	6	statistical	statistical	ADJ
esrj-89750	110	7	model	model	NOUN
esrj-89750	110	8	and	and	CCONJ
esrj-89750	110	9	physical	physical	ADJ
esrj-89750	110	10	model	model	NOUN
esrj-89750	110	11	according	accord	VERB
esrj-89750	110	12	to	to	ADP
esrj-89750	110	13	the	the	DET
esrj-89750	110	14	correction	correction	NOUN
esrj-89750	110	15	principle	principle	NOUN
esrj-89750	110	16	.	.	PUNCT
esrj-89750	111	1	the	the	DET
esrj-89750	111	2	statistical	statistical	ADJ
esrj-89750	111	3	model	model	NOUN
esrj-89750	111	4	is	be	AUX
esrj-89750	111	5	based	base	VERB
esrj-89750	111	6	on	on	ADP
esrj-89750	111	7	the	the	DET
esrj-89750	111	8	correlation	correlation	NOUN
esrj-89750	111	9	between	between	ADP
esrj-89750	111	10	surface	surface	NOUN
esrj-89750	111	11	variables	variable	NOUN
esrj-89750	111	12	and	and	CCONJ
esrj-89750	111	13	remote	remote	ADJ
esrj-89750	111	14	sensing	sense	VERB
esrj-89750	111	15	data	datum	NOUN
esrj-89750	111	16	,	,	PUNCT
esrj-89750	111	17	without	without	ADP
esrj-89750	111	18	knowing	know	VERB
esrj-89750	111	19	the	the	DET
esrj-89750	111	20	atmospheric	atmospheric	ADJ
esrj-89750	111	21	and	and	CCONJ
esrj-89750	111	22	geometric	geometric	ADJ
esrj-89750	111	23	conditions	condition	NOUN
esrj-89750	111	24	of	of	ADP
esrj-89750	111	25	image	image	NOUN
esrj-89750	111	26	acquisition	acquisition	NOUN
esrj-89750	111	27	.	.	PUNCT
esrj-89750	112	1	it	it	PRON
esrj-89750	112	2	has	have	VERB
esrj-89750	112	3	the	the	DET
esrj-89750	112	4	advantages	advantage	NOUN
esrj-89750	112	5	of	of	ADP
esrj-89750	112	6	simplicity	simplicity	NOUN
esrj-89750	112	7	and	and	CCONJ
esrj-89750	112	8	less	less	ADJ
esrj-89750	112	9	parameters	parameter	NOUN
esrj-89750	112	10	.	.	PUNCT
esrj-89750	113	1	2	2	X
esrj-89750	113	2	)	)	PUNCT
esrj-89750	113	3	geometric	geometric	ADJ
esrj-89750	113	4	distortion	distortion	NOUN
esrj-89750	113	5	correction	correction	NOUN
esrj-89750	113	6	.	.	PUNCT
esrj-89750	114	1	geometric	geometric	ADJ
esrj-89750	114	2	correction	correction	NOUN
esrj-89750	114	3	refers	refer	VERB
esrj-89750	114	4	to	to	ADP
esrj-89750	114	5	the	the	DET
esrj-89750	114	6	elimination	elimination	NOUN
esrj-89750	114	7	or	or	CCONJ
esrj-89750	114	8	correction	correction	NOUN
esrj-89750	114	9	of	of	ADP
esrj-89750	114	10	geometric	geometric	ADJ
esrj-89750	114	11	errors	error	NOUN
esrj-89750	114	12	in	in	ADP
esrj-89750	114	13	remote	remote	ADJ
esrj-89750	114	14	sensing	sensing	NOUN
esrj-89750	114	15	images	image	NOUN
esrj-89750	114	16	,	,	PUNCT
esrj-89750	114	17	which	which	PRON
esrj-89750	114	18	mainly	mainly	ADV
esrj-89750	114	19	involves	involve	VERB
esrj-89750	114	20	three	three	NUM
esrj-89750	114	21	processes	process	NOUN
esrj-89750	114	22	:	:	PUNCT
esrj-89750	114	23	selection	selection	NOUN
esrj-89750	114	24	of	of	ADP
esrj-89750	114	25	control	control	NOUN
esrj-89750	114	26	points	point	NOUN
esrj-89750	114	27	,	,	PUNCT
esrj-89750	114	28	transformation	transformation	NOUN
esrj-89750	114	29	of	of	ADP
esrj-89750	114	30	spatial	spatial	ADJ
esrj-89750	114	31	position	position	NOUN
esrj-89750	114	32	(	(	PUNCT
esrj-89750	114	33	coordinate	coordinate	NOUN
esrj-89750	114	34	transformation	transformation	NOUN
esrj-89750	114	35	)	)	PUNCT
esrj-89750	114	36	and	and	CCONJ
esrj-89750	114	37	resampling	resample	VERB
esrj-89750	114	38	of	of	ADP
esrj-89750	114	39	pixel	pixel	PROPN
esrj-89750	114	40	luminance	luminance	NOUN
esrj-89750	114	41	resampling	resample	VERB
esrj-89750	114	42	(	(	PUNCT
esrj-89750	114	43	pritikin	pritikin	NOUN
esrj-89750	114	44	et	et	PROPN
esrj-89750	114	45	al	al	PROPN
esrj-89750	114	46	.	.	PROPN
esrj-89750	114	47	,	,	PUNCT
esrj-89750	114	48	2018	2018	NUM
esrj-89750	114	49	)	)	PUNCT
esrj-89750	114	50	.	.	PUNCT
esrj-89750	115	1	control	control	NOUN
esrj-89750	115	2	point	point	NOUN
esrj-89750	115	3	selection	selection	NOUN
esrj-89750	115	4	:	:	PUNCT
esrj-89750	115	5	firstly	firstly	ADV
esrj-89750	115	6	,	,	PUNCT
esrj-89750	115	7	two	two	NUM
esrj-89750	115	8	remote	remote	ADJ
esrj-89750	115	9	sensing	sensing	NOUN
esrj-89750	115	10	images	image	NOUN
esrj-89750	115	11	are	be	AUX
esrj-89750	115	12	selected	select	VERB
esrj-89750	115	13	,	,	PUNCT
esrj-89750	115	14	one	one	NUM
esrj-89750	115	15	is	be	AUX
esrj-89750	115	16	the	the	DET
esrj-89750	115	17	reference	reference	NOUN
esrj-89750	115	18	image	image	NOUN
esrj-89750	115	19	and	and	CCONJ
esrj-89750	115	20	the	the	DET
esrj-89750	115	21	other	other	ADJ
esrj-89750	115	22	is	be	AUX
esrj-89750	115	23	the	the	DET
esrj-89750	115	24	geometric	geometric	ADJ
esrj-89750	115	25	distortion	distortion	NOUN
esrj-89750	115	26	image	image	NOUN
esrj-89750	115	27	,	,	PUNCT
esrj-89750	115	28	as	as	SCONJ
esrj-89750	115	29	shown	show	VERB
esrj-89750	115	30	in	in	ADP
esrj-89750	115	31	figure	figure	NOUN
esrj-89750	115	32	5	5	NUM
esrj-89750	115	33	.	.	PUNCT
esrj-89750	116	1	a	a	DET
esrj-89750	116	2	certain	certain	ADJ
esrj-89750	116	3	number	number	NOUN
esrj-89750	116	4	of	of	ADP
esrj-89750	116	5	control	control	NOUN
esrj-89750	116	6	point	point	NOUN
esrj-89750	116	7	pairs	pair	NOUN
esrj-89750	116	8	on	on	ADP
esrj-89750	116	9	the	the	DET
esrj-89750	116	10	above	above	ADJ
esrj-89750	116	11	two	two	NUM
esrj-89750	116	12	images	image	NOUN
esrj-89750	116	13	are	be	AUX
esrj-89750	116	14	selected	select	VERB
esrj-89750	116	15	.	.	PUNCT
esrj-89750	117	1	the	the	DET
esrj-89750	117	2	selection	selection	NOUN
esrj-89750	117	3	principles	principle	NOUN
esrj-89750	117	4	are	be	AUX
esrj-89750	117	5	as	as	SCONJ
esrj-89750	117	6	follows	follow	VERB
esrj-89750	117	7	:	:	PUNCT
esrj-89750	117	8	the	the	DET
esrj-89750	117	9	selected	select	VERB
esrj-89750	117	10	control	control	NOUN
esrj-89750	117	11	points	point	NOUN
esrj-89750	117	12	should	should	AUX
esrj-89750	117	13	be	be	AUX
esrj-89750	117	14	obvious	obvious	ADJ
esrj-89750	117	15	in	in	ADP
esrj-89750	117	16	the	the	DET
esrj-89750	117	17	remote	remote	ADJ
esrj-89750	117	18	sensing	sense	VERB
esrj-89750	117	19	image	image	NOUN
esrj-89750	117	20	of	of	ADP
esrj-89750	117	21	soil	soil	NOUN
esrj-89750	117	22	;	;	PUNCT
esrj-89750	117	23	the	the	DET
esrj-89750	117	24	objects	object	NOUN
esrj-89750	117	25	covered	cover	VERB
esrj-89750	117	26	by	by	ADP
esrj-89750	117	27	the	the	DET
esrj-89750	117	28	control	control	NOUN
esrj-89750	117	29	points	point	NOUN
esrj-89750	117	30	are	be	AUX
esrj-89750	117	31	always	always	ADV
esrj-89750	117	32	fixed	fix	VERB
esrj-89750	117	33	;	;	PUNCT
esrj-89750	117	34	the	the	DET
esrj-89750	117	35	selected	select	VERB
esrj-89750	117	36	control	control	NOUN
esrj-89750	117	37	points	point	NOUN
esrj-89750	117	38	have	have	VERB
esrj-89750	117	39	the	the	DET
esrj-89750	117	40	same	same	ADJ
esrj-89750	117	41	topographic	topographic	ADJ
esrj-89750	117	42	height	height	NOUN
esrj-89750	117	43	on	on	ADP
esrj-89750	117	44	the	the	DET
esrj-89750	117	45	two	two	NUM
esrj-89750	117	46	images	image	NOUN
esrj-89750	117	47	;	;	PUNCT
esrj-89750	117	48	the	the	DET
esrj-89750	117	49	control	control	NOUN
esrj-89750	117	50	points	point	NOUN
esrj-89750	117	51	should	should	AUX
esrj-89750	117	52	be	be	AUX
esrj-89750	117	53	evenly	evenly	ADV
esrj-89750	117	54	distributed	distribute	VERB
esrj-89750	117	55	in	in	ADP
esrj-89750	117	56	the	the	DET
esrj-89750	117	57	image	image	NOUN
esrj-89750	117	58	;	;	PUNCT
esrj-89750	117	59	and	and	CCONJ
esrj-89750	117	60	the	the	DET
esrj-89750	117	61	number	number	NOUN
esrj-89750	117	62	of	of	ADP
esrj-89750	117	63	control	control	NOUN
esrj-89750	117	64	points	point	NOUN
esrj-89750	117	65	should	should	AUX
esrj-89750	117	66	not	not	PART
esrj-89750	117	67	be	be	AUX
esrj-89750	117	68	less	less	ADJ
esrj-89750	117	69	than	than	ADP
esrj-89750	117	70	5	5	NUM
esrj-89750	117	71	.	.	PUNCT
esrj-89750	118	1	(	(	PUNCT
esrj-89750	118	2	a	a	X
esrj-89750	118	3	)	)	PUNCT
esrj-89750	118	4	base	base	NOUN
esrj-89750	118	5	image	image	NOUN
esrj-89750	118	6	(	(	PUNCT
esrj-89750	118	7	b	b	NOUN
esrj-89750	118	8	)	)	PUNCT
esrj-89750	118	9	distorted	distort	VERB
esrj-89750	118	10	image	image	NOUN
esrj-89750	118	11	figure	figure	NOUN
esrj-89750	118	12	5	5	NUM
esrj-89750	118	13	.	.	PUNCT
esrj-89750	118	14	control	control	NOUN
esrj-89750	118	15	point	point	NOUN
esrj-89750	118	16	selection	selection	NOUN
esrj-89750	118	17	space	space	NOUN
esrj-89750	118	18	position	position	NOUN
esrj-89750	118	19	transformation	transformation	NOUN
esrj-89750	118	20	(	(	PUNCT
esrj-89750	118	21	coordinate	coordinate	NOUN
esrj-89750	118	22	transformation	transformation	NOUN
esrj-89750	118	23	):	):	PUNCT
esrj-89750	118	24	the	the	DET
esrj-89750	118	25	purpose	purpose	NOUN
esrj-89750	118	26	is	be	AUX
esrj-89750	118	27	to	to	PART
esrj-89750	118	28	find	find	VERB
esrj-89750	118	29	the	the	DET
esrj-89750	118	30	correct	correct	ADJ
esrj-89750	118	31	coordinates	coordinate	NOUN
esrj-89750	118	32	of	of	ADP
esrj-89750	118	33	the	the	DET
esrj-89750	118	34	object	object	NOUN
esrj-89750	118	35	.	.	PUNCT
esrj-89750	119	1	there	there	PRON
esrj-89750	119	2	are	be	VERB
esrj-89750	119	3	two	two	NUM
esrj-89750	119	4	main	main	ADJ
esrj-89750	119	5	methods	method	NOUN
esrj-89750	119	6	at	at	ADP
esrj-89750	119	7	present	present	ADJ
esrj-89750	119	8	,	,	PUNCT
esrj-89750	119	9	direct	direct	ADJ
esrj-89750	119	10	method	method	NOUN
esrj-89750	119	11	and	and	CCONJ
esrj-89750	119	12	indirect	indirect	ADJ
esrj-89750	119	13	method	method	NOUN
esrj-89750	119	14	.	.	PUNCT
esrj-89750	120	1	direct	direct	ADJ
esrj-89750	120	2	method	method	NOUN
esrj-89750	120	3	,	,	PUNCT
esrj-89750	120	4	as	as	SCONJ
esrj-89750	120	5	its	its	PRON
esrj-89750	120	6	name	name	NOUN
esrj-89750	120	7	implies	imply	VERB
esrj-89750	120	8	,	,	PUNCT
esrj-89750	120	9	calculates	calculate	VERB
esrj-89750	120	10	the	the	DET
esrj-89750	120	11	coordinates	coordinate	NOUN
esrj-89750	120	12	of	of	ADP
esrj-89750	120	13	each	each	DET
esrj-89750	120	14	pixel	pixel	NOUN
esrj-89750	120	15	in	in	ADP
esrj-89750	120	16	the	the	DET
esrj-89750	120	17	image	image	NOUN
esrj-89750	120	18	to	to	PART
esrj-89750	120	19	be	be	AUX
esrj-89750	120	20	corrected	correct	VERB
esrj-89750	120	21	in	in	ADP
esrj-89750	120	22	turn	turn	NOUN
esrj-89750	120	23	by	by	ADP
esrj-89750	120	24	the	the	DET
esrj-89750	120	25	coordinates	coordinate	NOUN
esrj-89750	120	26	of	of	ADP
esrj-89750	120	27	the	the	DET
esrj-89750	120	28	control	control	NOUN
esrj-89750	120	29	points	point	NOUN
esrj-89750	120	30	on	on	ADP
esrj-89750	120	31	the	the	DET
esrj-89750	120	32	reference	reference	NOUN
esrj-89750	120	33	image	image	NOUN
esrj-89750	120	34	,	,	PUNCT
esrj-89750	120	35	as	as	SCONJ
esrj-89750	120	36	shown	show	VERB
esrj-89750	120	37	in	in	ADP
esrj-89750	120	38	figure	figure	NOUN
esrj-89750	120	39	6	6	NUM
esrj-89750	120	40	.	.	PUNCT
esrj-89750	121	1	(	(	PUNCT
esrj-89750	121	2	a	a	X
esrj-89750	121	3	)	)	PUNCT
esrj-89750	121	4	base	base	NOUN
esrj-89750	121	5	image	image	NOUN
esrj-89750	121	6	(	(	PUNCT
esrj-89750	121	7	b	b	NOUN
esrj-89750	121	8	)	)	PUNCT
esrj-89750	121	9	distorted	distort	VERB
esrj-89750	121	10	image	image	NOUN
esrj-89750	121	11	figure	figure	NOUN
esrj-89750	121	12	6	6	NUM
esrj-89750	121	13	.	.	PUNCT
esrj-89750	122	1	direct	direct	ADJ
esrj-89750	122	2	method	method	NOUN
esrj-89750	122	3	the	the	DET
esrj-89750	122	4	advantage	advantage	NOUN
esrj-89750	122	5	of	of	ADP
esrj-89750	122	6	the	the	DET
esrj-89750	122	7	direct	direct	ADJ
esrj-89750	122	8	method	method	NOUN
esrj-89750	122	9	is	be	AUX
esrj-89750	122	10	that	that	SCONJ
esrj-89750	122	11	the	the	DET
esrj-89750	122	12	coordinate	coordinate	ADJ
esrj-89750	122	13	values	value	NOUN
esrj-89750	122	14	of	of	ADP
esrj-89750	122	15	each	each	DET
esrj-89750	122	16	pixel	pixel	NOUN
esrj-89750	122	17	calculated	calculate	VERB
esrj-89750	122	18	in	in	ADP
esrj-89750	122	19	the	the	DET
esrj-89750	122	20	image	image	NOUN
esrj-89750	122	21	to	to	PART
esrj-89750	122	22	be	be	AUX
esrj-89750	122	23	corrected	correct	VERB
esrj-89750	122	24	will	will	AUX
esrj-89750	122	25	not	not	PART
esrj-89750	122	26	change	change	VERB
esrj-89750	122	27	.	.	PUNCT
esrj-89750	123	1	the	the	DET
esrj-89750	123	2	disadvantage	disadvantage	NOUN
esrj-89750	123	3	of	of	ADP
esrj-89750	123	4	the	the	DET
esrj-89750	123	5	direct	direct	ADJ
esrj-89750	123	6	method	method	NOUN
esrj-89750	123	7	is	be	AUX
esrj-89750	123	8	that	that	SCONJ
esrj-89750	123	9	the	the	DET
esrj-89750	123	10	distribution	distribution	NOUN
esrj-89750	123	11	of	of	ADP
esrj-89750	123	12	pixels	pixel	NOUN
esrj-89750	123	13	will	will	AUX
esrj-89750	123	14	not	not	PART
esrj-89750	123	15	be	be	AUX
esrj-89750	123	16	uniform	uniform	ADJ
esrj-89750	123	17	.	.	PUNCT
esrj-89750	124	1	in	in	ADP
esrj-89750	124	2	contrast	contrast	NOUN
esrj-89750	124	3	to	to	ADP
esrj-89750	124	4	the	the	DET
esrj-89750	124	5	direct	direct	ADJ
esrj-89750	124	6	method	method	NOUN
esrj-89750	124	7	,	,	PUNCT
esrj-89750	124	8	the	the	DET
esrj-89750	124	9	indirect	indirect	ADJ
esrj-89750	124	10	method	method	NOUN
esrj-89750	124	11	calculates	calculate	VERB
esrj-89750	124	12	the	the	DET
esrj-89750	124	13	coordinates	coordinate	NOUN
esrj-89750	124	14	of	of	ADP
esrj-89750	124	15	each	each	DET
esrj-89750	124	16	pixel	pixel	NOUN
esrj-89750	124	17	on	on	ADP
esrj-89750	124	18	the	the	DET
esrj-89750	124	19	reference	reference	NOUN
esrj-89750	124	20	image	image	NOUN
esrj-89750	124	21	from	from	ADP
esrj-89750	124	22	the	the	DET
esrj-89750	124	23	image	image	NOUN
esrj-89750	124	24	to	to	PART
esrj-89750	124	25	be	be	AUX
esrj-89750	124	26	corrected	correct	VERB
esrj-89750	124	27	.	.	PUNCT
esrj-89750	125	1	the	the	DET
esrj-89750	125	2	schematic	schematic	ADJ
esrj-89750	125	3	diagram	diagram	NOUN
esrj-89750	125	4	is	be	AUX
esrj-89750	125	5	shown	show	VERB
esrj-89750	125	6	in	in	ADP
esrj-89750	125	7	figure	figure	NOUN
esrj-89750	125	8	7	7	NUM
esrj-89750	125	9	.	.	PUNCT
esrj-89750	126	1	(	(	PUNCT
esrj-89750	126	2	a	a	DET
esrj-89750	126	3	)	)	PUNCT
esrj-89750	126	4	base	base	NOUN
esrj-89750	126	5	image	image	NOUN
esrj-89750	126	6	(	(	PUNCT
esrj-89750	126	7	b	b	NOUN
esrj-89750	126	8	)	)	PUNCT
esrj-89750	126	9	distorted	distort	VERB
esrj-89750	126	10	image	image	NOUN
esrj-89750	126	11	figure	figure	NOUN
esrj-89750	126	12	7	7	NUM
esrj-89750	126	13	.	.	PUNCT
esrj-89750	126	14	indirect	indirect	ADJ
esrj-89750	126	15	method	method	NOUN
esrj-89750	126	16	the	the	DET
esrj-89750	126	17	advantage	advantage	NOUN
esrj-89750	126	18	of	of	ADP
esrj-89750	126	19	the	the	DET
esrj-89750	126	20	indirect	indirect	ADJ
esrj-89750	126	21	method	method	NOUN
esrj-89750	126	22	is	be	AUX
esrj-89750	126	23	that	that	SCONJ
esrj-89750	126	24	it	it	PRON
esrj-89750	126	25	can	can	AUX
esrj-89750	126	26	ensure	ensure	VERB
esrj-89750	126	27	the	the	DET
esrj-89750	126	28	uniform	uniform	ADJ
esrj-89750	126	29	distribution	distribution	NOUN
esrj-89750	126	30	of	of	ADP
esrj-89750	126	31	the	the	DET
esrj-89750	126	32	corrected	correct	VERB
esrj-89750	126	33	image	image	NOUN
esrj-89750	126	34	pixels	pixel	NOUN
esrj-89750	126	35	in	in	ADP
esrj-89750	126	36	space	space	NOUN
esrj-89750	126	37	(	(	PUNCT
esrj-89750	126	38	temmer	temmer	NOUN
esrj-89750	126	39	et	et	PROPN
esrj-89750	126	40	al	al	PROPN
esrj-89750	126	41	.	.	PROPN
esrj-89750	126	42	,	,	PUNCT
esrj-89750	126	43	2017	2017	NUM
esrj-89750	126	44	)	)	PUNCT
esrj-89750	126	45	.	.	PUNCT
esrj-89750	127	1	the	the	DET
esrj-89750	127	2	disadvantage	disadvantage	NOUN
esrj-89750	127	3	of	of	ADP
esrj-89750	127	4	the	the	DET
esrj-89750	127	5	indirect	indirect	ADJ
esrj-89750	127	6	method	method	NOUN
esrj-89750	127	7	is	be	AUX
esrj-89750	127	8	that	that	SCONJ
esrj-89750	127	9	the	the	DET
esrj-89750	127	10	row	row	NOUN
esrj-89750	127	11	number	number	NOUN
esrj-89750	127	12	of	of	ADP
esrj-89750	127	13	the	the	DET
esrj-89750	127	14	relocated	relocate	VERB
esrj-89750	127	15	pixels	pixel	NOUN
esrj-89750	127	16	is	be	AUX
esrj-89750	127	17	not	not	PART
esrj-89750	127	18	an	an	DET
esrj-89750	127	19	integer	integer	NOUN
esrj-89750	127	20	relationship	relationship	NOUN
esrj-89750	127	21	with	with	ADP
esrj-89750	127	22	the	the	DET
esrj-89750	127	23	original	original	ADJ
esrj-89750	127	24	image	image	NOUN
esrj-89750	127	25	,	,	PUNCT
esrj-89750	127	26	so	so	CCONJ
esrj-89750	127	27	the	the	DET
esrj-89750	127	28	pixel	pixel	PROPN
esrj-89750	127	29	value	value	NOUN
esrj-89750	127	30	of	of	ADP
esrj-89750	127	31	the	the	DET
esrj-89750	127	32	original	original	ADJ
esrj-89750	127	33	image	image	NOUN
esrj-89750	127	34	needs	need	VERB
esrj-89750	127	35	to	to	PART
esrj-89750	127	36	be	be	AUX
esrj-89750	127	37	re	re	VERB
esrj-89750	127	38	-	-	VERB
esrj-89750	127	39	sampled	sample	VERB
esrj-89750	127	40	.	.	PUNCT
esrj-89750	128	1	pixel	pixel	PROPN
esrj-89750	128	2	luminance	luminance	PROPN
esrj-89750	128	3	resampling	resampling	NOUN
esrj-89750	128	4	:	:	PUNCT
esrj-89750	128	5	resampling	resample	VERB
esrj-89750	128	6	refers	refer	VERB
esrj-89750	128	7	to	to	ADP
esrj-89750	128	8	assigning	assign	VERB
esrj-89750	128	9	the	the	DET
esrj-89750	128	10	pixel	pixel	PROPN
esrj-89750	128	11	value	value	NOUN
esrj-89750	128	12	of	of	ADP
esrj-89750	128	13	the	the	DET
esrj-89750	128	14	original	original	ADJ
esrj-89750	128	15	image	image	NOUN
esrj-89750	128	16	to	to	ADP
esrj-89750	128	17	the	the	DET
esrj-89750	128	18	corrected	correct	VERB
esrj-89750	128	19	image	image	NOUN
esrj-89750	128	20	according	accord	VERB
esrj-89750	128	21	to	to	ADP
esrj-89750	128	22	a	a	DET
esrj-89750	128	23	certain	certain	ADJ
esrj-89750	128	24	relationship	relationship	NOUN
esrj-89750	128	25	.	.	PUNCT
esrj-89750	129	1	at	at	ADP
esrj-89750	129	2	present	present	ADJ
esrj-89750	129	3	,	,	PUNCT
esrj-89750	129	4	there	there	PRON
esrj-89750	129	5	are	be	VERB
esrj-89750	129	6	three	three	NUM
esrj-89750	129	7	main	main	ADJ
esrj-89750	129	8	methods	method	NOUN
esrj-89750	129	9	for	for	ADP
esrj-89750	129	10	resampling	resample	VERB
esrj-89750	129	11	pixel	pixel	PROPN
esrj-89750	129	12	luminance	luminance	NOUN
esrj-89750	129	13	values	value	NOUN
esrj-89750	129	14	:	:	PUNCT
esrj-89750	129	15	nearest	near	ADJ
esrj-89750	129	16	neighbor	neighbor	NOUN
esrj-89750	129	17	method	method	NOUN
esrj-89750	129	18	,	,	PUNCT
esrj-89750	129	19	bilinear	bilinear	ADJ
esrj-89750	129	20	interpolation	interpolation	NOUN
esrj-89750	129	21	method	method	NOUN
esrj-89750	129	22	and	and	CCONJ
esrj-89750	129	23	cubic	cubic	ADJ
esrj-89750	129	24	convolution	convolution	NOUN
esrj-89750	129	25	method	method	NOUN
esrj-89750	129	26	,	,	PUNCT
esrj-89750	129	27	as	as	SCONJ
esrj-89750	129	28	shown	show	VERB
esrj-89750	129	29	in	in	ADP
esrj-89750	129	30	table	table	NOUN
esrj-89750	129	31	2	2	NUM
esrj-89750	129	32	.	.	PUNCT
esrj-89750	129	33	target	target	NOUN
esrj-89750	129	34	detection	detection	NOUN
esrj-89750	129	35	of	of	ADP
esrj-89750	129	36	soil	soil	NOUN
esrj-89750	129	37	remote	remote	ADJ
esrj-89750	129	38	sensing	sense	VERB
esrj-89750	129	39	image	image	NOUN
esrj-89750	129	40	based	base	VERB
esrj-89750	129	41	on	on	ADP
esrj-89750	129	42	deep	deep	ADJ
esrj-89750	129	43	learning	learning	NOUN
esrj-89750	129	44	on	on	ADP
esrj-89750	129	45	the	the	DET
esrj-89750	129	46	basis	basis	NOUN
esrj-89750	129	47	of	of	ADP
esrj-89750	129	48	the	the	DET
esrj-89750	129	49	above	above	ADJ
esrj-89750	129	50	,	,	PUNCT
esrj-89750	129	51	this	this	DET
esrj-89750	129	52	chapter	chapter	NOUN
esrj-89750	129	53	uses	use	VERB
esrj-89750	129	54	a	a	DET
esrj-89750	129	55	deep	deep	ADJ
esrj-89750	129	56	learning	learning	NOUN
esrj-89750	129	57	algorithm	algorithm	NOUN
esrj-89750	129	58	to	to	PART
esrj-89750	129	59	detect	detect	VERB
esrj-89750	129	60	the	the	DET
esrj-89750	129	61	target	target	NOUN
esrj-89750	129	62	of	of	ADP
esrj-89750	129	63	soil	soil	NOUN
esrj-89750	129	64	remote	remote	ADJ
esrj-89750	129	65	sensing	sense	VERB
esrj-89750	129	66	image	image	NOUN
esrj-89750	129	67	,	,	PUNCT
esrj-89750	129	68	extract	extract	VERB
esrj-89750	129	69	soil	soil	NOUN
esrj-89750	129	70	characteristics	characteristic	NOUN
esrj-89750	129	71	,	,	PUNCT
esrj-89750	129	72	and	and	CCONJ
esrj-89750	129	73	prepare	prepare	VERB
esrj-89750	129	74	for	for	ADP
esrj-89750	129	75	subsequent	subsequent	ADJ
esrj-89750	129	76	classification	classification	NOUN
esrj-89750	129	77	.	.	PUNCT
esrj-89750	130	1	deep	deep	ADJ
esrj-89750	130	2	learning	learning	NOUN
esrj-89750	130	3	belongs	belong	VERB
esrj-89750	130	4	to	to	ADP
esrj-89750	130	5	a	a	DET
esrj-89750	130	6	branch	branch	NOUN
esrj-89750	130	7	of	of	ADP
esrj-89750	130	8	machine	machine	NOUN
esrj-89750	130	9	learning	learning	NOUN
esrj-89750	130	10	.	.	PUNCT
esrj-89750	131	1	it	it	PRON
esrj-89750	131	2	is	be	AUX
esrj-89750	131	3	an	an	DET
esrj-89750	131	4	algorithm	algorithm	NOUN
esrj-89750	131	5	based	base	VERB
esrj-89750	131	6	on	on	ADP
esrj-89750	131	7	an	an	DET
esrj-89750	131	8	artificial	artificial	ADJ
esrj-89750	131	9	neural	neural	ADJ
esrj-89750	131	10	network	network	NOUN
esrj-89750	131	11	to	to	PART
esrj-89750	131	12	represent	represent	VERB
esrj-89750	131	13	data	datum	NOUN
esrj-89750	131	14	.	.	PUNCT
esrj-89750	132	1	therefore	therefore	ADV
esrj-89750	132	2	,	,	PUNCT
esrj-89750	132	3	before	before	ADP
esrj-89750	132	4	analyzing	analyze	VERB
esrj-89750	132	5	deep	deep	ADJ
esrj-89750	132	6	learning	learning	NOUN
esrj-89750	132	7	,	,	PUNCT
esrj-89750	132	8	it	it	PRON
esrj-89750	132	9	is	be	AUX
esrj-89750	132	10	necessary	necessary	ADJ
esrj-89750	132	11	to	to	PART
esrj-89750	132	12	understand	understand	VERB
esrj-89750	132	13	artificial	artificial	ADJ
esrj-89750	132	14	neural	neural	ADJ
esrj-89750	132	15	networks	network	NOUN
esrj-89750	132	16	(	(	PUNCT
esrj-89750	132	17	wang	wang	PROPN
esrj-89750	132	18	et	et	PROPN
esrj-89750	132	19	al	al	PROPN
esrj-89750	132	20	.	.	PROPN
esrj-89750	132	21	,	,	PUNCT
esrj-89750	132	22	2017	2017	NUM
esrj-89750	132	23	)	)	PUNCT
esrj-89750	132	24	.	.	PUNCT
esrj-89750	133	1	artificial	artificial	ADJ
esrj-89750	133	2	neural	neural	ADJ
esrj-89750	133	3	network	network	NOUN
esrj-89750	133	4	(	(	PUNCT
esrj-89750	133	5	ann	ann	PROPN
esrj-89750	133	6	)	)	PUNCT
esrj-89750	133	7	is	be	AUX
esrj-89750	133	8	an	an	DET
esrj-89750	133	9	abstract	abstract	ADJ
esrj-89750	133	10	arithmetic	arithmetic	ADJ
esrj-89750	133	11	model	model	NOUN
esrj-89750	133	12	developed	develop	VERB
esrj-89750	133	13	to	to	PART
esrj-89750	133	14	simulate	simulate	VERB
esrj-89750	133	15	the	the	DET
esrj-89750	133	16	process	process	NOUN
esrj-89750	133	17	of	of	ADP
esrj-89750	133	18	processing	processing	NOUN
esrj-89750	133	19	information	information	NOUN
esrj-89750	133	20	by	by	ADP
esrj-89750	133	21	human	human	ADJ
esrj-89750	133	22	brain	brain	NOUN
esrj-89750	133	23	neurons	neuron	NOUN
esrj-89750	133	24	.	.	PUNCT
esrj-89750	134	1	it	it	PRON
esrj-89750	134	2	consists	consist	VERB
esrj-89750	134	3	of	of	ADP
esrj-89750	134	4	a	a	DET
esrj-89750	134	5	large	large	ADJ
esrj-89750	134	6	number	number	NOUN
esrj-89750	134	7	of	of	ADP
esrj-89750	134	8	nodes	node	NOUN
esrj-89750	134	9	(	(	PUNCT
esrj-89750	134	10	or	or	CCONJ
esrj-89750	134	11	neurons	neuron	NOUN
esrj-89750	134	12	)	)	PUNCT
esrj-89750	134	13	connected	connect	VERB
esrj-89750	134	14	with	with	ADP
esrj-89750	134	15	each	each	DET
esrj-89750	134	16	other	other	ADJ
esrj-89750	134	17	,	,	PUNCT
esrj-89750	134	18	as	as	SCONJ
esrj-89750	134	19	shown	show	VERB
esrj-89750	134	20	in	in	ADP
esrj-89750	134	21	figure	figure	NOUN
esrj-89750	134	22	8	8	NUM
esrj-89750	134	23	.	.	PUNCT
esrj-89750	135	1	in	in	ADP
esrj-89750	135	2	figure	figure	NOUN
esrj-89750	135	3	8	8	NUM
esrj-89750	135	4	,	,	PUNCT
esrj-89750	135	5	a	a	PRON
esrj-89750	135	6	is	be	AUX
esrj-89750	135	7	a	a	DET
esrj-89750	135	8	signal	signal	NOUN
esrj-89750	135	9	other	other	ADJ
esrj-89750	135	10	than	than	ADP
esrj-89750	135	11	the	the	DET
esrj-89750	135	12	neuron	neuron	NOUN
esrj-89750	135	13	;	;	PUNCT
esrj-89750	135	14	w1	w1	NOUN
esrj-89750	135	15	to	to	ADP
esrj-89750	135	16	wn	wn	PROPN
esrj-89750	135	17	is	be	AUX
esrj-89750	135	18	the	the	DET
esrj-89750	135	19	weight	weight	NOUN
esrj-89750	135	20	of	of	ADP
esrj-89750	135	21	the	the	DET
esrj-89750	135	22	signal	signal	NOUN
esrj-89750	135	23	transmitted	transmit	VERB
esrj-89750	135	24	to	to	ADP
esrj-89750	135	25	the	the	DET
esrj-89750	135	26	neuron	neuron	NOUN
esrj-89750	135	27	;	;	PUNCT
esrj-89750	135	28	b	b	X
esrj-89750	135	29	is	be	AUX
esrj-89750	135	30	the	the	DET
esrj-89750	135	31	bias	bias	NOUN
esrj-89750	135	32	of	of	ADP
esrj-89750	135	33	the	the	DET
esrj-89750	135	34	signal	signal	NOUN
esrj-89750	135	35	;	;	PUNCT
esrj-89750	135	36	f	f	PROPN
esrj-89750	135	37	is	be	AUX
esrj-89750	135	38	the	the	DET
esrj-89750	135	39	excitation	excitation	NOUN
esrj-89750	135	40	function	function	NOUN
esrj-89750	135	41	,	,	PUNCT
esrj-89750	135	42	generally	generally	ADV
esrj-89750	135	43	a	a	DET
esrj-89750	135	44	non	non	ADJ
esrj-89750	135	45	-	-	ADJ
esrj-89750	135	46	linear	linear	ADJ
esrj-89750	135	47	excitation	excitation	NOUN
esrj-89750	135	48	function	function	NOUN
esrj-89750	135	49	such	such	ADJ
esrj-89750	135	50	as	as	ADP
esrj-89750	135	51	sgn	sgn	ADJ
esrj-89750	135	52	symbolic	symbolic	ADJ
esrj-89750	135	53	function	function	NOUN
esrj-89750	135	54	or	or	CCONJ
esrj-89750	135	55	sigmoid	sigmoid	NOUN
esrj-89750	135	56	type	type	NOUN
esrj-89750	135	57	continuous	continuous	ADJ
esrj-89750	135	58	function	function	NOUN
esrj-89750	135	59	,	,	PUNCT
esrj-89750	135	60	and	and	CCONJ
esrj-89750	135	61	y	y	PROPN
esrj-89750	135	62	is	be	AUX
esrj-89750	135	63	the	the	DET
esrj-89750	135	64	output	output	NOUN
esrj-89750	135	65	of	of	ADP
esrj-89750	135	66	the	the	DET
esrj-89750	135	67	neuron	neuron	NOUN
esrj-89750	135	68	.	.	PUNCT
esrj-89750	136	1	artificial	artificial	ADJ
esrj-89750	136	2	neural	neural	ADJ
esrj-89750	136	3	networks	network	NOUN
esrj-89750	136	4	consist	consist	VERB
esrj-89750	136	5	of	of	ADP
esrj-89750	136	6	an	an	DET
esrj-89750	136	7	input	input	NOUN
esrj-89750	136	8	layer	layer	NOUN
esrj-89750	136	9	,	,	PUNCT
esrj-89750	136	10	hidden	hide	VERB
esrj-89750	136	11	layer	layer	NOUN
esrj-89750	136	12	and	and	CCONJ
esrj-89750	136	13	output	output	NOUN
esrj-89750	136	14	layer	layer	NOUN
esrj-89750	136	15	.	.	PUNCT
esrj-89750	137	1	when	when	SCONJ
esrj-89750	137	2	input	input	NOUN
esrj-89750	137	3	layer	layer	NOUN
esrj-89750	137	4	neurons	neuron	NOUN
esrj-89750	137	5	are	be	AUX
esrj-89750	137	6	stimulated	stimulate	VERB
esrj-89750	137	7	by	by	ADP
esrj-89750	137	8	input	input	NOUN
esrj-89750	137	9	signal	signal	NOUN
esrj-89750	137	10	,	,	PUNCT
esrj-89750	137	11	the	the	DET
esrj-89750	137	12	activation	activation	NOUN
esrj-89750	137	13	function	function	NOUN
esrj-89750	137	14	of	of	ADP
esrj-89750	137	15	neurons	neuron	NOUN
esrj-89750	137	16	in	in	ADP
esrj-89750	137	17	the	the	DET
esrj-89750	137	18	hidden	hide	VERB
esrj-89750	137	19	layer	layer	NOUN
esrj-89750	137	20	is	be	AUX
esrj-89750	137	21	stimulated	stimulate	VERB
esrj-89750	137	22	.	.	PUNCT
esrj-89750	138	1	when	when	SCONJ
esrj-89750	138	2	a	a	DET
esrj-89750	138	3	certain	certain	ADJ
esrj-89750	138	4	threshold	threshold	NOUN
esrj-89750	138	5	is	be	AUX
esrj-89750	138	6	reached	reach	VERB
esrj-89750	138	7	,	,	PUNCT
esrj-89750	138	8	neurons	neuron	NOUN
esrj-89750	138	9	are	be	AUX
esrj-89750	138	10	activated	activate	VERB
esrj-89750	138	11	and	and	CCONJ
esrj-89750	138	12	output	output	NOUN
esrj-89750	138	13	signal	signal	NOUN
esrj-89750	138	14	is	be	AUX
esrj-89750	138	15	generated	generate	VERB
esrj-89750	138	16	through	through	ADP
esrj-89750	138	17	the	the	DET
esrj-89750	138	18	output	output	NOUN
esrj-89750	138	19	layer	layer	NOUN
esrj-89750	138	20	(	(	PUNCT
esrj-89750	138	21	xu	xu	INTJ
esrj-89750	138	22	et	et	PROPN
esrj-89750	138	23	al	al	PROPN
esrj-89750	138	24	.	.	PROPN
esrj-89750	138	25	,	,	PUNCT
esrj-89750	138	26	2017	2017	NUM
esrj-89750	138	27	)	)	PUNCT
esrj-89750	138	28	.	.	PUNCT
esrj-89750	139	1	on	on	ADP
esrj-89750	139	2	the	the	DET
esrj-89750	139	3	basis	basis	NOUN
esrj-89750	139	4	of	of	ADP
esrj-89750	139	5	this	this	DET
esrj-89750	139	6	structure	structure	NOUN
esrj-89750	139	7	,	,	PUNCT
esrj-89750	139	8	it	it	PRON
esrj-89750	139	9	is	be	AUX
esrj-89750	139	10	proposed	propose	VERB
esrj-89750	139	11	that	that	SCONJ
esrj-89750	139	12	deep	deep	ADJ
esrj-89750	139	13	learning	learning	NOUN
esrj-89750	139	14	is	be	AUX
esrj-89750	139	15	a	a	DET
esrj-89750	139	16	deeper	deep	ADJ
esrj-89750	139	17	artificial	artificial	ADJ
esrj-89750	139	18	neural	neural	ADJ
esrj-89750	139	19	network	network	NOUN
esrj-89750	139	20	,	,	PUNCT
esrj-89750	139	21	which	which	PRON
esrj-89750	139	22	consists	consist	VERB
esrj-89750	139	23	of	of	ADP
esrj-89750	139	24	multiple	multiple	ADJ
esrj-89750	139	25	neurons	neuron	NOUN
esrj-89750	139	26	stacked	stack	VERB
esrj-89750	139	27	together	together	ADV
esrj-89750	139	28	,	,	PUNCT
esrj-89750	139	29	as	as	SCONJ
esrj-89750	139	30	shown	show	VERB
esrj-89750	139	31	in	in	ADP
esrj-89750	139	32	figure	figure	NOUN
esrj-89750	139	33	9	9	NUM
esrj-89750	139	34	.	.	PUNCT
esrj-89750	140	1	there	there	PRON
esrj-89750	140	2	is	be	VERB
esrj-89750	140	3	a	a	DET
esrj-89750	140	4	special	special	ADJ
esrj-89750	140	5	network	network	NOUN
esrj-89750	140	6	structure	structure	NOUN
esrj-89750	140	7	in	in	ADP
esrj-89750	140	8	deep	deep	ADJ
esrj-89750	140	9	learning	learning	NOUN
esrj-89750	140	10	,	,	PUNCT
esrj-89750	140	11	namely	namely	ADV
esrj-89750	140	12	convolution	convolution	VERB
esrj-89750	140	13	neural	neural	ADJ
esrj-89750	140	14	networks	network	NOUN
esrj-89750	140	15	.	.	PUNCT
esrj-89750	141	1	it	it	PRON
esrj-89750	141	2	is	be	AUX
esrj-89750	141	3	a	a	DET
esrj-89750	141	4	kind	kind	NOUN
esrj-89750	141	5	of	of	ADP
esrj-89750	141	6	feedforward	feedforward	ADJ
esrj-89750	141	7	neural	neural	ADJ
esrj-89750	141	8	network	network	NOUN
esrj-89750	141	9	with	with	ADP
esrj-89750	141	10	deep	deep	ADJ
esrj-89750	141	11	structure	structure	NOUN
esrj-89750	141	12	including	include	VERB
esrj-89750	141	13	convolution	convolution	NOUN
esrj-89750	141	14	calculation	calculation	NOUN
esrj-89750	141	15	(	(	PUNCT
esrj-89750	141	16	zhao	zhao	PROPN
esrj-89750	141	17	et	et	PROPN
esrj-89750	141	18	al	al	PROPN
esrj-89750	141	19	.	.	PROPN
esrj-89750	141	20	,	,	PUNCT
esrj-89750	141	21	2018	2018	NUM
esrj-89750	141	22	)	)	PUNCT
esrj-89750	141	23	.	.	PUNCT
esrj-89750	142	1	its	its	PRON
esrj-89750	142	2	structure	structure	NOUN
esrj-89750	142	3	consists	consist	VERB
esrj-89750	142	4	of	of	ADP
esrj-89750	142	5	input	input	NOUN
esrj-89750	142	6	layer	layer	NOUN
esrj-89750	142	7	,	,	PUNCT
esrj-89750	142	8	convolution	convolution	NOUN
esrj-89750	142	9	layer	layer	NOUN
esrj-89750	142	10	,	,	PUNCT
esrj-89750	142	11	pool	pool	NOUN
esrj-89750	142	12	layer	layer	NOUN
esrj-89750	142	13	,	,	PUNCT
esrj-89750	142	14	full	full	ADJ
esrj-89750	142	15	connection	connection	NOUN
esrj-89750	142	16	layer	layer	NOUN
esrj-89750	142	17	and	and	CCONJ
esrj-89750	142	18	output	output	NOUN
esrj-89750	142	19	layer	layer	NOUN
esrj-89750	142	20	.	.	PUNCT
esrj-89750	143	1	the	the	DET
esrj-89750	143	2	structure	structure	NOUN
esrj-89750	143	3	is	be	AUX
esrj-89750	143	4	shown	show	VERB
esrj-89750	143	5	in	in	ADP
esrj-89750	143	6	figure	figure	NOUN
esrj-89750	143	7	10	10	NUM
esrj-89750	143	8	.	.	PUNCT
esrj-89750	144	1	the	the	DET
esrj-89750	144	2	basic	basic	ADJ
esrj-89750	144	3	principle	principle	NOUN
esrj-89750	144	4	of	of	ADP
esrj-89750	144	5	convolution	convolution	NOUN
esrj-89750	144	6	neural	neural	ADJ
esrj-89750	144	7	networks	network	NOUN
esrj-89750	144	8	is	be	AUX
esrj-89750	144	9	that	that	SCONJ
esrj-89750	144	10	the	the	DET
esrj-89750	144	11	input	input	NOUN
esrj-89750	144	12	image	image	NOUN
esrj-89750	144	13	is	be	AUX
esrj-89750	144	14	convoluted	convolute	VERB
esrj-89750	144	15	by	by	ADP
esrj-89750	144	16	three	three	NUM
esrj-89750	144	17	trainable	trainable	ADJ
esrj-89750	144	18	filters	filter	NOUN
esrj-89750	144	19	and	and	CCONJ
esrj-89750	144	20	additive	additive	ADJ
esrj-89750	144	21	bias	bias	NOUN
esrj-89750	144	22	.	.	PUNCT
esrj-89750	145	1	after	after	ADP
esrj-89750	145	2	convolution	convolution	NOUN
esrj-89750	145	3	,	,	PUNCT
esrj-89750	145	4	three	three	NUM
esrj-89750	145	5	feature	feature	NOUN
esrj-89750	145	6	mapping	mapping	NOUN
esrj-89750	145	7	maps	map	NOUN
esrj-89750	145	8	are	be	AUX
esrj-89750	145	9	generated	generate	VERB
esrj-89750	145	10	at	at	ADP
esrj-89750	145	11	layer	layer	NOUN
esrj-89750	145	12	c1	c1	PROPN
esrj-89750	145	13	.	.	PUNCT
esrj-89750	146	1	then	then	ADV
esrj-89750	146	2	four	four	NUM
esrj-89750	146	3	pixels	pixel	NOUN
esrj-89750	146	4	of	of	ADP
esrj-89750	146	5	each	each	DET
esrj-89750	146	6	group	group	NOUN
esrj-89750	146	7	in	in	ADP
esrj-89750	146	8	the	the	DET
esrj-89750	146	9	feature	feature	NOUN
esrj-89750	146	10	mapping	mapping	NOUN
esrj-89750	146	11	map	map	NOUN
esrj-89750	146	12	are	be	AUX
esrj-89750	146	13	summated	summate	VERB
esrj-89750	146	14	,	,	PUNCT
esrj-89750	146	15	weighted	weighted	ADJ
esrj-89750	146	16	and	and	CCONJ
esrj-89750	146	17	biased	bias	VERB
esrj-89750	146	18	,	,	PUNCT
esrj-89750	146	19	and	and	CCONJ
esrj-89750	146	20	three	three	NUM
esrj-89750	146	21	feature	feature	NOUN
esrj-89750	146	22	mapping	mapping	NOUN
esrj-89750	146	23	maps	map	NOUN
esrj-89750	146	24	at	at	ADP
esrj-89750	146	25	layer	layer	NOUN
esrj-89750	146	26	s2	s2	PROPN
esrj-89750	146	27	are	be	AUX
esrj-89750	146	28	obtained	obtain	VERB
esrj-89750	146	29	through	through	ADP
esrj-89750	146	30	a	a	DET
esrj-89750	146	31	sigmoid	sigmoid	NOUN
esrj-89750	146	32	function	function	NOUN
esrj-89750	146	33	.	.	PUNCT
esrj-89750	147	1	361maximum	361maximum	NUM
esrj-89750	147	2	likelihood	likelihood	NOUN
esrj-89750	147	3	classification	classification	NOUN
esrj-89750	147	4	of	of	ADP
esrj-89750	147	5	soil	soil	NOUN
esrj-89750	147	6	remote	remote	ADJ
esrj-89750	147	7	sensing	sense	VERB
esrj-89750	147	8	image	image	NOUN
esrj-89750	147	9	based	base	VERB
esrj-89750	147	10	on	on	ADP
esrj-89750	147	11	deep	deep	ADJ
esrj-89750	147	12	learning	learn	VERB
esrj-89750	147	13	these	these	DET
esrj-89750	147	14	maps	map	NOUN
esrj-89750	147	15	are	be	AUX
esrj-89750	147	16	filtered	filter	VERB
esrj-89750	147	17	to	to	PART
esrj-89750	147	18	get	get	VERB
esrj-89750	147	19	layer	layer	NOUN
esrj-89750	147	20	c3	c3	PROPN
esrj-89750	147	21	.	.	PUNCT
esrj-89750	148	1	this	this	DET
esrj-89750	148	2	hierarchy	hierarchy	NOUN
esrj-89750	148	3	produces	produce	VERB
esrj-89750	148	4	layer	layer	NOUN
esrj-89750	148	5	s4	s4	NOUN
esrj-89750	148	6	again	again	ADV
esrj-89750	148	7	,	,	PUNCT
esrj-89750	148	8	as	as	SCONJ
esrj-89750	148	9	does	do	VERB
esrj-89750	148	10	layer	layer	NOUN
esrj-89750	148	11	s2	s2	PROPN
esrj-89750	148	12	.	.	PUNCT
esrj-89750	149	1	finally	finally	ADV
esrj-89750	149	2	,	,	PUNCT
esrj-89750	149	3	these	these	DET
esrj-89750	149	4	pixel	pixel	PROPN
esrj-89750	149	5	values	value	NOUN
esrj-89750	149	6	are	be	AUX
esrj-89750	149	7	rasterized	rasterize	VERB
esrj-89750	149	8	and	and	CCONJ
esrj-89750	149	9	connected	connect	VERB
esrj-89750	149	10	into	into	ADP
esrj-89750	149	11	a	a	DET
esrj-89750	149	12	vector	vector	NOUN
esrj-89750	149	13	input	input	NOUN
esrj-89750	149	14	to	to	ADP
esrj-89750	149	15	the	the	DET
esrj-89750	149	16	traditional	traditional	ADJ
esrj-89750	149	17	neural	neural	ADJ
esrj-89750	149	18	network	network	NOUN
esrj-89750	149	19	to	to	PART
esrj-89750	149	20	get	get	VERB
esrj-89750	149	21	the	the	DET
esrj-89750	149	22	output	output	NOUN
esrj-89750	149	23	(	(	PUNCT
esrj-89750	149	24	zhu	zhu	X
esrj-89750	149	25	et	et	PROPN
esrj-89750	149	26	al	al	PROPN
esrj-89750	149	27	.	.	PROPN
esrj-89750	149	28	,	,	PUNCT
esrj-89750	149	29	2019	2019	NUM
esrj-89750	149	30	)	)	PUNCT
esrj-89750	149	31	.	.	PUNCT
esrj-89750	150	1	based	base	VERB
esrj-89750	150	2	on	on	ADP
esrj-89750	150	3	the	the	DET
esrj-89750	150	4	above	above	ADJ
esrj-89750	150	5	convolution	convolution	NOUN
esrj-89750	150	6	neural	neural	ADJ
esrj-89750	150	7	network	network	NOUN
esrj-89750	150	8	,	,	PUNCT
esrj-89750	150	9	the	the	DET
esrj-89750	150	10	object	object	NOUN
esrj-89750	150	11	detection	detection	NOUN
esrj-89750	150	12	and	and	CCONJ
esrj-89750	150	13	extraction	extraction	NOUN
esrj-89750	150	14	of	of	ADP
esrj-89750	150	15	soil	soil	NOUN
esrj-89750	150	16	remote	remote	ADJ
esrj-89750	150	17	sensing	sensing	NOUN
esrj-89750	150	18	images	image	NOUN
esrj-89750	150	19	are	be	AUX
esrj-89750	150	20	carried	carry	VERB
esrj-89750	150	21	out	out	ADP
esrj-89750	150	22	.	.	PUNCT
esrj-89750	151	1	the	the	DET
esrj-89750	151	2	specific	specific	ADJ
esrj-89750	151	3	process	process	NOUN
esrj-89750	151	4	is	be	AUX
esrj-89750	151	5	as	as	SCONJ
esrj-89750	151	6	follows	follow	VERB
esrj-89750	151	7	:	:	PUNCT
esrj-89750	151	8	(	(	PUNCT
esrj-89750	151	9	1	1	X
esrj-89750	151	10	)	)	PUNCT
esrj-89750	151	11	convolutional	convolutional	ADJ
esrj-89750	151	12	neural	neural	ADJ
esrj-89750	151	13	network	network	NOUN
esrj-89750	151	14	training	training	NOUN
esrj-89750	151	15	step	step	NOUN
esrj-89750	151	16	1	1	NUM
esrj-89750	151	17	:	:	PUNCT
esrj-89750	151	18	select	select	VERB
esrj-89750	151	19	the	the	DET
esrj-89750	151	20	remote	remote	ADJ
esrj-89750	151	21	sensing	sense	VERB
esrj-89750	151	22	image	image	NOUN
esrj-89750	151	23	of	of	ADP
esrj-89750	151	24	soil	soil	NOUN
esrj-89750	151	25	for	for	ADP
esrj-89750	151	26	training	training	NOUN
esrj-89750	151	27	as	as	ADP
esrj-89750	151	28	the	the	DET
esrj-89750	151	29	input	input	NOUN
esrj-89750	151	30	sample	sample	NOUN
esrj-89750	151	31	of	of	ADP
esrj-89750	151	32	the	the	DET
esrj-89750	151	33	model	model	NOUN
esrj-89750	151	34	.	.	PUNCT
esrj-89750	152	1	step	step	NOUN
esrj-89750	152	2	2	2	NUM
esrj-89750	152	3	:	:	PUNCT
esrj-89750	152	4	set	set	VERB
esrj-89750	152	5	the	the	DET
esrj-89750	152	6	weights	weight	NOUN
esrj-89750	152	7	between	between	ADP
esrj-89750	152	8	layers	layer	NOUN
esrj-89750	152	9	,	,	PUNCT
esrj-89750	152	10	the	the	DET
esrj-89750	152	11	thresholds	threshold	NOUN
esrj-89750	152	12	of	of	ADP
esrj-89750	152	13	output	output	NOUN
esrj-89750	152	14	units	unit	NOUN
esrj-89750	152	15	and	and	CCONJ
esrj-89750	152	16	hidden	hidden	ADJ
esrj-89750	152	17	units	unit	NOUN
esrj-89750	152	18	to	to	ADP
esrj-89750	152	19	random	random	ADJ
esrj-89750	152	20	values	value	NOUN
esrj-89750	152	21	close	close	ADV
esrj-89750	152	22	to	to	ADP
esrj-89750	152	23	0	0	NUM
esrj-89750	152	24	,	,	PUNCT
esrj-89750	152	25	and	and	CCONJ
esrj-89750	152	26	initialize	initialize	VERB
esrj-89750	152	27	the	the	DET
esrj-89750	152	28	accuracy	accuracy	NOUN
esrj-89750	152	29	control	control	NOUN
esrj-89750	152	30	parameters	parameter	NOUN
esrj-89750	152	31	and	and	CCONJ
esrj-89750	152	32	learning	learn	VERB
esrj-89750	152	33	rate	rate	NOUN
esrj-89750	152	34	of	of	ADP
esrj-89750	152	35	the	the	DET
esrj-89750	152	36	model	model	NOUN
esrj-89750	152	37	.	.	PUNCT
esrj-89750	153	1	step	step	NOUN
esrj-89750	153	2	3	3	NUM
esrj-89750	153	3	:	:	PUNCT
esrj-89750	153	4	take	take	VERB
esrj-89750	153	5	an	an	DET
esrj-89750	153	6	input	input	NOUN
esrj-89750	153	7	mode	mode	NOUN
esrj-89750	153	8	x	x	PUNCT
esrj-89750	153	9	from	from	ADP
esrj-89750	153	10	the	the	DET
esrj-89750	153	11	training	training	NOUN
esrj-89750	153	12	group	group	NOUN
esrj-89750	153	13	and	and	CCONJ
esrj-89750	153	14	add	add	VERB
esrj-89750	153	15	it	it	PRON
esrj-89750	153	16	to	to	ADP
esrj-89750	153	17	the	the	DET
esrj-89750	153	18	network	network	NOUN
esrj-89750	153	19	,	,	PUNCT
esrj-89750	153	20	and	and	CCONJ
esrj-89750	153	21	give	give	VERB
esrj-89750	153	22	its	its	PRON
esrj-89750	153	23	target	target	NOUN
esrj-89750	153	24	output	output	NOUN
esrj-89750	153	25	vector	vector	NOUN
esrj-89750	153	26	d.	d.	PROPN
esrj-89750	153	27	step	step	NOUN
esrj-89750	153	28	4	4	NUM
esrj-89750	153	29	:	:	PUNCT
esrj-89750	153	30	calculate	calculate	VERB
esrj-89750	153	31	an	an	DET
esrj-89750	153	32	intermediate	intermediate	ADJ
esrj-89750	153	33	output	output	NOUN
esrj-89750	153	34	vector	vector	NOUN
esrj-89750	153	35	h	h	NOUN
esrj-89750	153	36	and	and	CCONJ
esrj-89750	153	37	the	the	DET
esrj-89750	153	38	actual	actual	ADJ
esrj-89750	153	39	output	output	NOUN
esrj-89750	153	40	vector	vector	NOUN
esrj-89750	153	41	y	y	PROPN
esrj-89750	153	42	of	of	ADP
esrj-89750	153	43	the	the	DET
esrj-89750	153	44	network	network	NOUN
esrj-89750	153	45	.	.	PUNCT
esrj-89750	154	1	step	step	NOUN
esrj-89750	154	2	5	5	NUM
esrj-89750	154	3	:	:	PUNCT
esrj-89750	154	4	compare	compare	VERB
esrj-89750	154	5	the	the	DET
esrj-89750	154	6	element	element	NOUN
esrj-89750	154	7	yk	yk	PROPN
esrj-89750	154	8	in	in	ADP
esrj-89750	154	9	the	the	DET
esrj-89750	154	10	output	output	NOUN
esrj-89750	154	11	vector	vector	NOUN
esrj-89750	154	12	with	with	ADP
esrj-89750	154	13	the	the	DET
esrj-89750	154	14	element	element	NOUN
esrj-89750	154	15	dk	dk	PROPN
esrj-89750	154	16	in	in	ADP
esrj-89750	154	17	the	the	DET
esrj-89750	154	18	target	target	NOUN
esrj-89750	154	19	vector	vector	NOUN
esrj-89750	154	20	,	,	PUNCT
esrj-89750	154	21	and	and	CCONJ
esrj-89750	154	22	calculate	calculate	NOUN
esrj-89750	154	23	m	m	NOUN
esrj-89750	154	24	output	output	NOUN
esrj-89750	154	25	error	error	NOUN
esrj-89750	154	26	terms	term	NOUN
esrj-89750	154	27	.	.	PUNCT
esrj-89750	155	1	table	table	NOUN
esrj-89750	155	2	2	2	NUM
esrj-89750	155	3	.	.	PUNCT
esrj-89750	155	4	resampling	resample	VERB
esrj-89750	155	5	method	method	NOUN
esrj-89750	155	6	of	of	ADP
esrj-89750	155	7	pixel	pixel	PROPN
esrj-89750	155	8	luminance	luminance	PROPN
esrj-89750	155	9	method	method	NOUN
esrj-89750	155	10	definition	definition	NOUN
esrj-89750	155	11	characteristic	characteristic	ADJ
esrj-89750	155	12	sketch	sketch	NOUN
esrj-89750	155	13	map	map	NOUN
esrj-89750	155	14	knn	knn	VERB
esrj-89750	155	15	the	the	DET
esrj-89750	155	16	gray	gray	ADJ
esrj-89750	155	17	value	value	NOUN
esrj-89750	155	18	of	of	ADP
esrj-89750	155	19	the	the	DET
esrj-89750	155	20	nearest	near	ADJ
esrj-89750	155	21	neighbor	neighbor	NOUN
esrj-89750	155	22	of	of	ADP
esrj-89750	155	23	the	the	DET
esrj-89750	155	24	four	four	NUM
esrj-89750	155	25	adjacent	adjacent	ADJ
esrj-89750	155	26	pixels	pixel	NOUN
esrj-89750	155	27	around	around	ADP
esrj-89750	155	28	the	the	DET
esrj-89750	155	29	sampling	sampling	NOUN
esrj-89750	155	30	point	point	NOUN
esrj-89750	155	31	is	be	AUX
esrj-89750	155	32	taken	take	VERB
esrj-89750	155	33	as	as	ADP
esrj-89750	155	34	the	the	DET
esrj-89750	155	35	gray	gray	ADJ
esrj-89750	155	36	value	value	NOUN
esrj-89750	155	37	of	of	ADP
esrj-89750	155	38	the	the	DET
esrj-89750	155	39	point	point	NOUN
esrj-89750	155	40	.	.	PUNCT
esrj-89750	156	1	the	the	DET
esrj-89750	156	2	method	method	NOUN
esrj-89750	156	3	is	be	AUX
esrj-89750	156	4	simple	simple	ADJ
esrj-89750	156	5	,	,	PUNCT
esrj-89750	156	6	fast	fast	ADJ
esrj-89750	156	7	and	and	CCONJ
esrj-89750	156	8	does	do	AUX
esrj-89750	156	9	not	not	PART
esrj-89750	156	10	change	change	VERB
esrj-89750	156	11	the	the	DET
esrj-89750	156	12	value	value	NOUN
esrj-89750	156	13	of	of	ADP
esrj-89750	156	14	the	the	DET
esrj-89750	156	15	original	original	ADJ
esrj-89750	156	16	raster	raster	NOUN
esrj-89750	156	17	,	,	PUNCT
esrj-89750	156	18	but	but	CCONJ
esrj-89750	156	19	the	the	DET
esrj-89750	156	20	processed	process	VERB
esrj-89750	156	21	image	image	NOUN
esrj-89750	156	22	is	be	AUX
esrj-89750	156	23	not	not	PART
esrj-89750	156	24	smooth	smooth	ADJ
esrj-89750	156	25	enough	enough	ADV
esrj-89750	156	26	.	.	PUNCT
esrj-89750	157	1	bilinear	bilinear	NOUN
esrj-89750	157	2	interpolation	interpolation	NOUN
esrj-89750	157	3	linear	linear	ADJ
esrj-89750	157	4	interpolation	interpolation	NOUN
esrj-89750	157	5	of	of	ADP
esrj-89750	157	6	gray	gray	ADJ
esrj-89750	157	7	values	value	NOUN
esrj-89750	157	8	of	of	ADP
esrj-89750	157	9	four	four	NUM
esrj-89750	157	10	adjacent	adjacent	ADJ
esrj-89750	157	11	points	point	NOUN
esrj-89750	157	12	in	in	ADP
esrj-89750	157	13	two	two	NUM
esrj-89750	157	14	directions	direction	NOUN
esrj-89750	157	15	is	be	AUX
esrj-89750	157	16	used	use	VERB
esrj-89750	157	17	to	to	PART
esrj-89750	157	18	obtain	obtain	VERB
esrj-89750	157	19	gray	gray	ADJ
esrj-89750	157	20	values	value	NOUN
esrj-89750	157	21	of	of	ADP
esrj-89750	157	22	points	point	NOUN
esrj-89750	157	23	to	to	PART
esrj-89750	157	24	be	be	AUX
esrj-89750	157	25	sampled	sample	VERB
esrj-89750	157	26	.	.	PUNCT
esrj-89750	158	1	the	the	DET
esrj-89750	158	2	result	result	NOUN
esrj-89750	158	3	is	be	AUX
esrj-89750	158	4	smoother	smooth	ADJ
esrj-89750	158	5	than	than	ADP
esrj-89750	158	6	that	that	PRON
esrj-89750	158	7	of	of	ADP
esrj-89750	158	8	the	the	DET
esrj-89750	158	9	nearest	near	ADJ
esrj-89750	158	10	neighbor	neighbor	NOUN
esrj-89750	158	11	method	method	NOUN
esrj-89750	158	12	,	,	PUNCT
esrj-89750	158	13	but	but	CCONJ
esrj-89750	158	14	it	it	PRON
esrj-89750	158	15	will	will	AUX
esrj-89750	158	16	change	change	VERB
esrj-89750	158	17	the	the	DET
esrj-89750	158	18	original	original	ADJ
esrj-89750	158	19	grid	grid	NOUN
esrj-89750	158	20	value	value	NOUN
esrj-89750	158	21	and	and	CCONJ
esrj-89750	158	22	lose	lose	VERB
esrj-89750	158	23	some	some	DET
esrj-89750	158	24	small	small	ADJ
esrj-89750	158	25	features	feature	NOUN
esrj-89750	158	26	.	.	PUNCT
esrj-89750	159	1	it	it	PRON
esrj-89750	159	2	is	be	AUX
esrj-89750	159	3	suitable	suitable	ADJ
esrj-89750	159	4	for	for	ADP
esrj-89750	159	5	continuous	continuous	ADJ
esrj-89750	159	6	data	datum	NOUN
esrj-89750	159	7	representing	represent	VERB
esrj-89750	159	8	the	the	DET
esrj-89750	159	9	distribution	distribution	NOUN
esrj-89750	159	10	of	of	ADP
esrj-89750	159	11	some	some	DET
esrj-89750	159	12	phenomena	phenomenon	NOUN
esrj-89750	159	13	and	and	CCONJ
esrj-89750	159	14	topographic	topographic	ADJ
esrj-89750	159	15	surface	surface	NOUN
esrj-89750	159	16	.	.	PUNCT
esrj-89750	160	1	cubic	cubic	ADJ
esrj-89750	160	2	convolution	convolution	NOUN
esrj-89750	160	3	interpolation	interpolation	NOUN
esrj-89750	160	4	cubic	cubic	ADJ
esrj-89750	160	5	interpolation	interpolation	NOUN
esrj-89750	160	6	using	use	VERB
esrj-89750	160	7	the	the	DET
esrj-89750	160	8	gray	gray	ADJ
esrj-89750	160	9	value	value	NOUN
esrj-89750	160	10	of	of	ADP
esrj-89750	160	11	the	the	DET
esrj-89750	160	12	pixels	pixel	NOUN
esrj-89750	160	13	in	in	ADP
esrj-89750	160	14	the	the	DET
esrj-89750	160	15	larger	large	ADJ
esrj-89750	160	16	neighborhood	neighborhood	NOUN
esrj-89750	160	17	around	around	ADP
esrj-89750	160	18	the	the	DET
esrj-89750	160	19	sampling	sampling	NOUN
esrj-89750	160	20	point	point	NOUN
esrj-89750	160	21	.	.	PUNCT
esrj-89750	161	1	it	it	PRON
esrj-89750	161	2	can	can	AUX
esrj-89750	161	3	make	make	VERB
esrj-89750	161	4	the	the	DET
esrj-89750	161	5	image	image	NOUN
esrj-89750	161	6	smooth	smooth	VERB
esrj-89750	161	7	and	and	CCONJ
esrj-89750	161	8	have	have	VERB
esrj-89750	161	9	a	a	DET
esrj-89750	161	10	good	good	ADJ
esrj-89750	161	11	visual	visual	ADJ
esrj-89750	161	12	effect	effect	NOUN
esrj-89750	161	13	,	,	PUNCT
esrj-89750	161	14	but	but	CCONJ
esrj-89750	161	15	it	it	PRON
esrj-89750	161	16	will	will	AUX
esrj-89750	161	17	destroy	destroy	VERB
esrj-89750	161	18	the	the	DET
esrj-89750	161	19	spectral	spectral	ADJ
esrj-89750	161	20	information	information	NOUN
esrj-89750	161	21	of	of	ADP
esrj-89750	161	22	the	the	DET
esrj-89750	161	23	image	image	NOUN
esrj-89750	161	24	.	.	PUNCT
esrj-89750	162	1	this	this	DET
esrj-89750	162	2	method	method	NOUN
esrj-89750	162	3	can	can	AUX
esrj-89750	162	4	be	be	AUX
esrj-89750	162	5	used	use	VERB
esrj-89750	162	6	when	when	SCONJ
esrj-89750	162	7	data	datum	NOUN
esrj-89750	162	8	processing	processing	NOUN
esrj-89750	162	9	based	base	VERB
esrj-89750	162	10	on	on	ADP
esrj-89750	162	11	spectral	spectral	ADJ
esrj-89750	162	12	analysis	analysis	NOUN
esrj-89750	162	13	is	be	AUX
esrj-89750	162	14	no	no	ADV
esrj-89750	162	15	longer	long	ADV
esrj-89750	162	16	needed	need	VERB
esrj-89750	162	17	,	,	PUNCT
esrj-89750	162	18	but	but	CCONJ
esrj-89750	162	19	only	only	ADV
esrj-89750	162	20	for	for	ADP
esrj-89750	162	21	graphical	graphical	ADJ
esrj-89750	162	22	representation	representation	NOUN
esrj-89750	162	23	.	.	PUNCT
esrj-89750	163	1	https://blog.csdn.net/u010510549/article/details/62427122	https://blog.csdn.net/u010510549/article/details/62427122	X
esrj-89750	163	2	https://blog.csdn.net/u010510549/article/details/62427122	https://blog.csdn.net/u010510549/article/details/62427122	NUM
esrj-89750	164	1	362	362	NUM
esrj-89750	164	2	shujun	shujun	PROPN
esrj-89750	164	3	liang1	liang1	PROPN
esrj-89750	164	4	,	,	PUNCT
esrj-89750	164	5	jing	je	VERB
esrj-89750	165	1	cheng2	cheng2	PROPN
esrj-89750	165	2	*	*	PROPN
esrj-89750	165	3	,	,	PUNCT
esrj-89750	165	4	jianwei	jianwei	PROPN
esrj-89750	165	5	zhang1	zhang1	NOUN
esrj-89750	165	6	step	step	NOUN
esrj-89750	165	7	6	6	NUM
esrj-89750	165	8	:	:	PUNCT
esrj-89750	165	9	calculate	calculate	NOUN
esrj-89750	165	10	l	l	NOUN
esrj-89750	165	11	error	error	NOUN
esrj-89750	165	12	terms	term	NOUN
esrj-89750	165	13	for	for	ADP
esrj-89750	165	14	hidden	hidden	ADJ
esrj-89750	165	15	elements	element	NOUN
esrj-89750	165	16	in	in	ADP
esrj-89750	165	17	the	the	DET
esrj-89750	165	18	middle	middle	ADJ
esrj-89750	165	19	layer	layer	NOUN
esrj-89750	165	20	.	.	PUNCT
esrj-89750	166	1	step	step	NOUN
esrj-89750	166	2	7	7	NUM
esrj-89750	166	3	:	:	PUNCT
esrj-89750	166	4	calculate	calculate	VERB
esrj-89750	166	5	the	the	DET
esrj-89750	166	6	adjustment	adjustment	NOUN
esrj-89750	166	7	formula	formula	NOUN
esrj-89750	166	8	of	of	ADP
esrj-89750	166	9	each	each	DET
esrj-89750	166	10	weight	weight	NOUN
esrj-89750	166	11	value	value	NOUN
esrj-89750	166	12	and	and	CCONJ
esrj-89750	166	13	the	the	DET
esrj-89750	166	14	adjustment	adjustment	NOUN
esrj-89750	166	15	formula	formula	NOUN
esrj-89750	166	16	of	of	ADP
esrj-89750	166	17	threshold	threshold	NOUN
esrj-89750	166	18	value	value	NOUN
esrj-89750	166	19	in	in	ADP
esrj-89750	166	20	turn	turn	NOUN
esrj-89750	166	21	.	.	PUNCT
esrj-89750	167	1	step	step	NOUN
esrj-89750	167	2	8	8	NUM
esrj-89750	167	3	:	:	PUNCT
esrj-89750	167	4	adjust	adjust	VERB
esrj-89750	167	5	weights	weight	NOUN
esrj-89750	167	6	and	and	CCONJ
esrj-89750	167	7	thresholds	threshold	NOUN
esrj-89750	167	8	.	.	PUNCT
esrj-89750	168	1	step	step	VERB
esrj-89750	168	2	9	9	NUM
esrj-89750	168	3	:	:	PUNCT
esrj-89750	168	4	when	when	SCONJ
esrj-89750	168	5	k	k	PROPN
esrj-89750	168	6	goes	go	VERB
esrj-89750	168	7	through	through	ADP
esrj-89750	168	8	1	1	NUM
esrj-89750	168	9	to	to	ADP
esrj-89750	168	10	m	m	PROPN
esrj-89750	168	11	,	,	PUNCT
esrj-89750	168	12	judge	judge	VERB
esrj-89750	168	13	whether	whether	SCONJ
esrj-89750	168	14	the	the	DET
esrj-89750	168	15	index	index	NOUN
esrj-89750	168	16	meets	meet	VERB
esrj-89750	168	17	the	the	DET
esrj-89750	168	18	accuracy	accuracy	NOUN
esrj-89750	168	19	requirement	requirement	NOUN
esrj-89750	168	20	:	:	PUNCT
esrj-89750	168	21	e≤ε	e≤ε	PROPN
esrj-89750	168	22	,	,	PUNCT
esrj-89750	168	23	where	where	SCONJ
esrj-89750	168	24	e	e	NOUN
esrj-89750	168	25	is	be	AUX
esrj-89750	168	26	the	the	DET
esrj-89750	168	27	total	total	ADJ
esrj-89750	168	28	error	error	NOUN
esrj-89750	168	29	function	function	NOUN
esrj-89750	168	30	.	.	PUNCT
esrj-89750	169	1	if	if	SCONJ
esrj-89750	169	2	not	not	PART
esrj-89750	169	3	satisfied	satisfied	ADJ
esrj-89750	169	4	,	,	PUNCT
esrj-89750	169	5	go	go	VERB
esrj-89750	169	6	back	back	ADV
esrj-89750	169	7	to	to	PART
esrj-89750	169	8	step	step	VERB
esrj-89750	169	9	3	3	NUM
esrj-89750	169	10	and	and	CCONJ
esrj-89750	169	11	continue	continue	VERB
esrj-89750	169	12	iterating	iterate	VERB
esrj-89750	169	13	;	;	PUNCT
esrj-89750	169	14	if	if	SCONJ
esrj-89750	169	15	satisfied	satisfied	ADJ
esrj-89750	169	16	,	,	PUNCT
esrj-89750	169	17	go	go	VERB
esrj-89750	169	18	to	to	ADP
esrj-89750	169	19	the	the	DET
esrj-89750	169	20	next	next	ADJ
esrj-89750	169	21	step	step	NOUN
esrj-89750	169	22	.	.	PUNCT
esrj-89750	170	1	step	step	NOUN
esrj-89750	170	2	10	10	NUM
esrj-89750	170	3	:	:	PUNCT
esrj-89750	171	1	at	at	ADP
esrj-89750	171	2	the	the	DET
esrj-89750	171	3	end	end	NOUN
esrj-89750	171	4	of	of	ADP
esrj-89750	171	5	the	the	DET
esrj-89750	171	6	training	training	NOUN
esrj-89750	171	7	,	,	PUNCT
esrj-89750	171	8	save	save	VERB
esrj-89750	171	9	the	the	DET
esrj-89750	171	10	weights	weight	NOUN
esrj-89750	171	11	and	and	CCONJ
esrj-89750	171	12	thresholds	threshold	NOUN
esrj-89750	171	13	in	in	ADP
esrj-89750	171	14	the	the	DET
esrj-89750	171	15	file	file	NOUN
esrj-89750	171	16	.	.	PUNCT
esrj-89750	172	1	at	at	ADP
esrj-89750	172	2	this	this	DET
esrj-89750	172	3	time	time	NOUN
esrj-89750	172	4	,	,	PUNCT
esrj-89750	172	5	it	it	PRON
esrj-89750	172	6	can	can	AUX
esrj-89750	172	7	be	be	AUX
esrj-89750	172	8	considered	consider	VERB
esrj-89750	172	9	that	that	SCONJ
esrj-89750	172	10	the	the	DET
esrj-89750	172	11	weights	weight	NOUN
esrj-89750	172	12	have	have	AUX
esrj-89750	172	13	been	be	AUX
esrj-89750	172	14	stabilized	stabilize	VERB
esrj-89750	172	15	and	and	CCONJ
esrj-89750	172	16	classifiers	classifier	NOUN
esrj-89750	172	17	have	have	AUX
esrj-89750	172	18	been	be	AUX
esrj-89750	172	19	formed	form	VERB
esrj-89750	172	20	.	.	PUNCT
esrj-89750	173	1	when	when	SCONJ
esrj-89750	173	2	training	train	VERB
esrj-89750	173	3	again	again	ADV
esrj-89750	173	4	,	,	PUNCT
esrj-89750	173	5	the	the	DET
esrj-89750	173	6	weights	weight	NOUN
esrj-89750	173	7	and	and	CCONJ
esrj-89750	173	8	thresholds	threshold	NOUN
esrj-89750	173	9	are	be	AUX
esrj-89750	173	10	directly	directly	ADV
esrj-89750	173	11	derived	derive	VERB
esrj-89750	173	12	from	from	ADP
esrj-89750	173	13	the	the	DET
esrj-89750	173	14	file	file	NOUN
esrj-89750	173	15	for	for	ADP
esrj-89750	173	16	training	training	NOUN
esrj-89750	173	17	without	without	ADP
esrj-89750	173	18	initialization	initialization	NOUN
esrj-89750	173	19	.	.	PUNCT
esrj-89750	174	1	(	(	PUNCT
esrj-89750	174	2	2	2	X
esrj-89750	174	3	)	)	PUNCT
esrj-89750	174	4	realization	realization	NOUN
esrj-89750	174	5	of	of	ADP
esrj-89750	174	6	soil	soil	NOUN
esrj-89750	174	7	remote	remote	ADJ
esrj-89750	174	8	sensing	sense	VERB
esrj-89750	174	9	image	image	NOUN
esrj-89750	174	10	detection	detection	NOUN
esrj-89750	174	11	step	step	NOUN
esrj-89750	174	12	1	1	NUM
esrj-89750	174	13	:	:	PUNCT
esrj-89750	174	14	use	use	VERB
esrj-89750	174	15	the	the	DET
esrj-89750	174	16	trained	train	VERB
esrj-89750	174	17	convolution	convolution	NOUN
esrj-89750	174	18	neural	neural	ADJ
esrj-89750	174	19	network	network	NOUN
esrj-89750	174	20	to	to	PART
esrj-89750	174	21	extract	extract	VERB
esrj-89750	174	22	image	image	NOUN
esrj-89750	174	23	texture	texture	NOUN
esrj-89750	174	24	features	feature	NOUN
esrj-89750	174	25	,	,	PUNCT
esrj-89750	174	26	and	and	CCONJ
esrj-89750	174	27	convolution	convolution	NOUN
esrj-89750	174	28	calculation	calculation	NOUN
esrj-89750	174	29	is	be	AUX
esrj-89750	174	30	carried	carry	VERB
esrj-89750	174	31	out	out	ADP
esrj-89750	174	32	to	to	PART
esrj-89750	174	33	obtain	obtain	VERB
esrj-89750	174	34	image	image	NOUN
esrj-89750	174	35	feature	feature	NOUN
esrj-89750	174	36	map	map	NOUN
esrj-89750	174	37	.	.	PUNCT
esrj-89750	175	1	step	step	NOUN
esrj-89750	175	2	2	2	NUM
esrj-89750	175	3	:	:	PUNCT
esrj-89750	175	4	make	make	VERB
esrj-89750	175	5	sampling	sampling	NOUN
esrj-89750	175	6	of	of	ADP
esrj-89750	175	7	the	the	DET
esrj-89750	175	8	feature	feature	NOUN
esrj-89750	175	9	map	map	NOUN
esrj-89750	175	10	of	of	ADP
esrj-89750	175	11	soil	soil	NOUN
esrj-89750	175	12	remote	remote	ADJ
esrj-89750	175	13	sensing	sense	VERB
esrj-89750	175	14	image	image	NOUN
esrj-89750	175	15	by	by	ADP
esrj-89750	175	16	using	use	VERB
esrj-89750	175	17	adaptive	adaptive	ADJ
esrj-89750	175	18	pooling	pooling	NOUN
esrj-89750	175	19	model	model	NOUN
esrj-89750	175	20	step	step	NOUN
esrj-89750	175	21	3	3	NUM
esrj-89750	175	22	:	:	PUNCT
esrj-89750	175	23	merge	merge	VERB
esrj-89750	175	24	the	the	DET
esrj-89750	175	25	feature	feature	NOUN
esrj-89750	175	26	map	map	NOUN
esrj-89750	175	27	into	into	ADP
esrj-89750	175	28	a	a	DET
esrj-89750	175	29	column	column	NOUN
esrj-89750	175	30	of	of	ADP
esrj-89750	175	31	feature	feature	NOUN
esrj-89750	175	32	vectors	vector	NOUN
esrj-89750	175	33	,	,	PUNCT
esrj-89750	175	34	and	and	CCONJ
esrj-89750	175	35	input	input	NOUN
esrj-89750	175	36	to	to	ADP
esrj-89750	175	37	the	the	DET
esrj-89750	175	38	full	full	ADJ
esrj-89750	175	39	connection	connection	NOUN
esrj-89750	175	40	layer	layer	NOUN
esrj-89750	175	41	.	.	PUNCT
esrj-89750	176	1	update	update	VERB
esrj-89750	176	2	the	the	DET
esrj-89750	176	3	weights	weight	NOUN
esrj-89750	176	4	of	of	ADP
esrj-89750	176	5	the	the	DET
esrj-89750	176	6	network	network	NOUN
esrj-89750	176	7	filter	filter	NOUN
esrj-89750	176	8	through	through	ADP
esrj-89750	176	9	label	label	NOUN
esrj-89750	176	10	data	datum	NOUN
esrj-89750	176	11	back	back	ADV
esrj-89750	176	12	propagation	propagation	NOUN
esrj-89750	176	13	algorithm	algorithm	NOUN
esrj-89750	176	14	.	.	PUNCT
esrj-89750	177	1	step	step	NOUN
esrj-89750	177	2	4	4	NUM
esrj-89750	177	3	:	:	PUNCT
esrj-89750	177	4	finally	finally	ADV
esrj-89750	177	5	,	,	PUNCT
esrj-89750	177	6	the	the	DET
esrj-89750	177	7	feature	feature	NOUN
esrj-89750	177	8	column	column	NOUN
esrj-89750	177	9	vectors	vector	NOUN
esrj-89750	177	10	are	be	AUX
esrj-89750	177	11	input	input	VERB
esrj-89750	177	12	into	into	ADP
esrj-89750	177	13	softmax	softmax	NOUN
esrj-89750	177	14	to	to	PART
esrj-89750	177	15	complete	complete	VERB
esrj-89750	177	16	the	the	DET
esrj-89750	177	17	object	object	NOUN
esrj-89750	177	18	extraction	extraction	NOUN
esrj-89750	177	19	of	of	ADP
esrj-89750	177	20	soil	soil	NOUN
esrj-89750	177	21	image	image	NOUN
esrj-89750	177	22	.	.	PUNCT
esrj-89750	178	1	figure	figure	NOUN
esrj-89750	178	2	8	8	NUM
esrj-89750	178	3	.	.	PUNCT
esrj-89750	179	1	artificial	artificial	ADJ
esrj-89750	179	2	neural	neural	ADJ
esrj-89750	179	3	network	network	NOUN
esrj-89750	179	4	figure	figure	NOUN
esrj-89750	179	5	9	9	NUM
esrj-89750	179	6	.	.	PUNCT
esrj-89750	180	1	deep	deep	ADJ
esrj-89750	180	2	artificial	artificial	ADJ
esrj-89750	180	3	neural	neural	ADJ
esrj-89750	180	4	network	network	NOUN
esrj-89750	180	5	figure	figure	NOUN
esrj-89750	180	6	10	10	NUM
esrj-89750	180	7	.	.	PUNCT
esrj-89750	181	1	convolutional	convolutional	ADJ
esrj-89750	181	2	neural	neural	ADJ
esrj-89750	181	3	network	network	NOUN
esrj-89750	181	4	363maximum	363maximum	PROPN
esrj-89750	181	5	likelihood	likelihood	NOUN
esrj-89750	181	6	classification	classification	NOUN
esrj-89750	181	7	of	of	ADP
esrj-89750	181	8	soil	soil	NOUN
esrj-89750	181	9	remote	remote	ADJ
esrj-89750	181	10	sensing	sense	VERB
esrj-89750	181	11	image	image	NOUN
esrj-89750	181	12	based	base	VERB
esrj-89750	181	13	on	on	ADP
esrj-89750	181	14	deep	deep	ADJ
esrj-89750	181	15	learning	learn	VERB
esrj-89750	181	16	classification	classification	NOUN
esrj-89750	181	17	of	of	ADP
esrj-89750	181	18	soil	soil	NOUN
esrj-89750	181	19	remote	remote	ADJ
esrj-89750	181	20	sensing	sense	VERB
esrj-89750	181	21	image	image	NOUN
esrj-89750	181	22	based	base	VERB
esrj-89750	181	23	on	on	ADP
esrj-89750	181	24	maximum	maximum	ADJ
esrj-89750	181	25	likelihood	likelihood	NOUN
esrj-89750	181	26	estimation	estimation	NOUN
esrj-89750	181	27	maximum	maximum	ADJ
esrj-89750	181	28	likelihood	likelihood	NOUN
esrj-89750	181	29	classification	classification	NOUN
esrj-89750	181	30	(	(	PUNCT
esrj-89750	181	31	mlc	mlc	PROPN
esrj-89750	181	32	)	)	PUNCT
esrj-89750	181	33	has	have	VERB
esrj-89750	181	34	a	a	DET
esrj-89750	181	35	rigorous	rigorous	ADJ
esrj-89750	181	36	theoretical	theoretical	ADJ
esrj-89750	181	37	basis	basis	NOUN
esrj-89750	181	38	.	.	PUNCT
esrj-89750	182	1	it	it	PRON
esrj-89750	182	2	is	be	AUX
esrj-89750	182	3	easy	easy	ADJ
esrj-89750	182	4	to	to	PART
esrj-89750	182	5	establish	establish	VERB
esrj-89750	182	6	a	a	DET
esrj-89750	182	7	class	class	NOUN
esrj-89750	182	8	discriminant	discriminant	NOUN
esrj-89750	182	9	function	function	VERB
esrj-89750	182	10	with	with	ADP
esrj-89750	182	11	normal	normal	ADJ
esrj-89750	182	12	distribution	distribution	NOUN
esrj-89750	182	13	.	.	PUNCT
esrj-89750	183	1	it	it	PRON
esrj-89750	183	2	combines	combine	VERB
esrj-89750	183	3	the	the	DET
esrj-89750	183	4	mean	mean	ADJ
esrj-89750	183	5	,	,	PUNCT
esrj-89750	183	6	variance	variance	NOUN
esrj-89750	183	7	and	and	CCONJ
esrj-89750	183	8	covariance	covariance	NOUN
esrj-89750	183	9	of	of	ADP
esrj-89750	183	10	each	each	DET
esrj-89750	183	11	category	category	NOUN
esrj-89750	183	12	in	in	ADP
esrj-89750	183	13	each	each	DET
esrj-89750	183	14	band	band	NOUN
esrj-89750	183	15	.	.	PUNCT
esrj-89750	184	1	it	it	PRON
esrj-89750	184	2	has	have	VERB
esrj-89750	184	3	good	good	ADJ
esrj-89750	184	4	statistical	statistical	ADJ
esrj-89750	184	5	characteristics	characteristic	NOUN
esrj-89750	184	6	and	and	CCONJ
esrj-89750	184	7	has	have	AUX
esrj-89750	184	8	been	be	AUX
esrj-89750	184	9	considered	consider	VERB
esrj-89750	184	10	as	as	ADP
esrj-89750	184	11	a	a	DET
esrj-89750	184	12	more	more	ADV
esrj-89750	184	13	advanced	advanced	ADJ
esrj-89750	184	14	classification	classification	NOUN
esrj-89750	184	15	method	method	NOUN
esrj-89750	184	16	.	.	PUNCT
esrj-89750	185	1	in	in	ADP
esrj-89750	185	2	traditional	traditional	ADJ
esrj-89750	185	3	remote	remote	ADJ
esrj-89750	185	4	sensing	sense	VERB
esrj-89750	185	5	image	image	NOUN
esrj-89750	185	6	classification	classification	NOUN
esrj-89750	185	7	,	,	PUNCT
esrj-89750	185	8	maximum	maximum	ADJ
esrj-89750	185	9	likelihood	likelihood	NOUN
esrj-89750	185	10	method	method	NOUN
esrj-89750	185	11	is	be	AUX
esrj-89750	185	12	widely	widely	ADV
esrj-89750	185	13	used	use	VERB
esrj-89750	185	14	.	.	PUNCT
esrj-89750	186	1	this	this	DET
esrj-89750	186	2	method	method	NOUN
esrj-89750	186	3	obtains	obtain	VERB
esrj-89750	186	4	the	the	DET
esrj-89750	186	5	mean	mean	NOUN
esrj-89750	186	6	and	and	CCONJ
esrj-89750	186	7	variance	variance	NOUN
esrj-89750	186	8	of	of	ADP
esrj-89750	186	9	each	each	DET
esrj-89750	186	10	category	category	NOUN
esrj-89750	186	11	by	by	ADP
esrj-89750	186	12	statistics	statistic	NOUN
esrj-89750	186	13	and	and	CCONJ
esrj-89750	186	14	calculation	calculation	NOUN
esrj-89750	186	15	of	of	ADP
esrj-89750	186	16	the	the	DET
esrj-89750	186	17	interested	interested	ADJ
esrj-89750	186	18	region	region	NOUN
esrj-89750	186	19	,	,	PUNCT
esrj-89750	186	20	and	and	CCONJ
esrj-89750	186	21	then	then	ADV
esrj-89750	186	22	determines	determine	VERB
esrj-89750	186	23	a	a	DET
esrj-89750	186	24	classification	classification	NOUN
esrj-89750	186	25	function	function	NOUN
esrj-89750	186	26	.	.	PUNCT
esrj-89750	187	1	then	then	ADV
esrj-89750	187	2	each	each	DET
esrj-89750	187	3	pixel	pixel	NOUN
esrj-89750	187	4	in	in	ADP
esrj-89750	187	5	the	the	DET
esrj-89750	187	6	image	image	NOUN
esrj-89750	187	7	to	to	PART
esrj-89750	187	8	be	be	AUX
esrj-89750	187	9	classified	classify	VERB
esrj-89750	187	10	is	be	AUX
esrj-89750	187	11	substituted	substitute	VERB
esrj-89750	187	12	into	into	ADP
esrj-89750	187	13	the	the	DET
esrj-89750	187	14	classification	classification	NOUN
esrj-89750	187	15	function	function	NOUN
esrj-89750	187	16	of	of	ADP
esrj-89750	187	17	each	each	DET
esrj-89750	187	18	category	category	NOUN
esrj-89750	187	19	,	,	PUNCT
esrj-89750	187	20	and	and	CCONJ
esrj-89750	187	21	the	the	DET
esrj-89750	187	22	category	category	NOUN
esrj-89750	187	23	with	with	ADP
esrj-89750	187	24	the	the	DET
esrj-89750	187	25	largest	large	ADJ
esrj-89750	187	26	return	return	NOUN
esrj-89750	187	27	value	value	NOUN
esrj-89750	187	28	of	of	ADP
esrj-89750	187	29	the	the	DET
esrj-89750	187	30	function	function	NOUN
esrj-89750	187	31	is	be	AUX
esrj-89750	187	32	regarded	regard	VERB
esrj-89750	187	33	as	as	ADP
esrj-89750	187	34	the	the	DET
esrj-89750	187	35	category	category	NOUN
esrj-89750	187	36	of	of	ADP
esrj-89750	187	37	the	the	DET
esrj-89750	187	38	scanned	scan	VERB
esrj-89750	187	39	pixels	pixel	NOUN
esrj-89750	187	40	,	,	PUNCT
esrj-89750	187	41	so	so	SCONJ
esrj-89750	187	42	as	as	SCONJ
esrj-89750	187	43	to	to	PART
esrj-89750	187	44	achieve	achieve	VERB
esrj-89750	187	45	the	the	DET
esrj-89750	187	46	classification	classification	NOUN
esrj-89750	187	47	effect	effect	NOUN
esrj-89750	187	48	.	.	PUNCT
esrj-89750	188	1	its	its	PRON
esrj-89750	188	2	basic	basic	ADJ
esrj-89750	188	3	principle	principle	NOUN
esrj-89750	188	4	process	process	NOUN
esrj-89750	188	5	is	be	AUX
esrj-89750	188	6	as	as	SCONJ
esrj-89750	188	7	follows	follow	VERB
esrj-89750	188	8	:	:	PUNCT
esrj-89750	188	9	step	step	NOUN
esrj-89750	188	10	1	1	NUM
esrj-89750	188	11	:	:	PUNCT
esrj-89750	188	12	determine	determine	VERB
esrj-89750	188	13	the	the	DET
esrj-89750	188	14	area	area	NOUN
esrj-89750	188	15	to	to	PART
esrj-89750	188	16	be	be	AUX
esrj-89750	188	17	classified	classify	VERB
esrj-89750	188	18	and	and	CCONJ
esrj-89750	188	19	the	the	DET
esrj-89750	188	20	number	number	NOUN
esrj-89750	188	21	of	of	ADP
esrj-89750	188	22	bands	band	NOUN
esrj-89750	188	23	and	and	CCONJ
esrj-89750	188	24	feature	feature	NOUN
esrj-89750	188	25	classifications	classification	NOUN
esrj-89750	188	26	to	to	PART
esrj-89750	188	27	be	be	AUX
esrj-89750	188	28	used	use	VERB
esrj-89750	188	29	,	,	PUNCT
esrj-89750	188	30	and	and	CCONJ
esrj-89750	188	31	check	check	VERB
esrj-89750	188	32	whether	whether	SCONJ
esrj-89750	188	33	each	each	DET
esrj-89750	188	34	band	band	NOUN
esrj-89750	188	35	or	or	CCONJ
esrj-89750	188	36	feature	feature	NOUN
esrj-89750	188	37	component	component	NOUN
esrj-89750	188	38	has	have	AUX
esrj-89750	188	39	been	be	AUX
esrj-89750	188	40	positioned	position	VERB
esrj-89750	188	41	with	with	ADP
esrj-89750	188	42	each	each	DET
esrj-89750	188	43	other	other	ADJ
esrj-89750	188	44	.	.	PUNCT
esrj-89750	189	1	step	step	NOUN
esrj-89750	189	2	2	2	NUM
esrj-89750	189	3	:	:	PUNCT
esrj-89750	189	4	according	accord	VERB
esrj-89750	189	5	to	to	ADP
esrj-89750	189	6	the	the	DET
esrj-89750	189	7	ground	ground	NOUN
esrj-89750	189	8	condition	condition	NOUN
esrj-89750	189	9	of	of	ADP
esrj-89750	189	10	the	the	DET
esrj-89750	189	11	typical	typical	ADJ
esrj-89750	189	12	area	area	NOUN
esrj-89750	189	13	,	,	PUNCT
esrj-89750	189	14	choose	choose	VERB
esrj-89750	189	15	the	the	DET
esrj-89750	189	16	training	training	NOUN
esrj-89750	189	17	area	area	NOUN
esrj-89750	189	18	on	on	ADP
esrj-89750	189	19	the	the	DET
esrj-89750	189	20	image	image	NOUN
esrj-89750	189	21	.	.	PUNCT
esrj-89750	190	1	step	step	NOUN
esrj-89750	190	2	3	3	NUM
esrj-89750	190	3	:	:	PUNCT
esrj-89750	190	4	calculation	calculation	NOUN
esrj-89750	190	5	of	of	ADP
esrj-89750	190	6	the	the	DET
esrj-89750	190	7	parameters	parameter	NOUN
esrj-89750	190	8	.	.	PUNCT
esrj-89750	191	1	calculate	calculate	VERB
esrj-89750	191	2	and	and	CCONJ
esrj-89750	191	3	determine	determine	VERB
esrj-89750	191	4	the	the	DET
esrj-89750	191	5	prior	prior	ADJ
esrj-89750	191	6	probability	probability	NOUN
esrj-89750	191	7	according	accord	VERB
esrj-89750	191	8	to	to	ADP
esrj-89750	191	9	the	the	DET
esrj-89750	191	10	image	image	NOUN
esrj-89750	191	11	data	datum	NOUN
esrj-89750	191	12	of	of	ADP
esrj-89750	191	13	the	the	DET
esrj-89750	191	14	selected	select	VERB
esrj-89750	191	15	training	training	NOUN
esrj-89750	191	16	areas	area	NOUN
esrj-89750	191	17	.	.	PUNCT
esrj-89750	192	1	step	step	VERB
esrj-89750	192	2	4	4	NUM
esrj-89750	192	3	:	:	PUNCT
esrj-89750	192	4	classification	classification	NOUN
esrj-89750	192	5	.	.	PUNCT
esrj-89750	193	1	substitute	substitute	VERB
esrj-89750	193	2	the	the	DET
esrj-89750	193	3	image	image	NOUN
esrj-89750	193	4	pixels	pixel	VERB
esrj-89750	193	5	outside	outside	ADP
esrj-89750	193	6	the	the	DET
esrj-89750	193	7	training	training	NOUN
esrj-89750	193	8	area	area	NOUN
esrj-89750	193	9	into	into	ADP
esrj-89750	193	10	the	the	DET
esrj-89750	193	11	formula	formula	NOUN
esrj-89750	193	12	one	one	NUM
esrj-89750	193	13	by	by	ADP
esrj-89750	193	14	one	one	NUM
esrj-89750	193	15	.	.	PUNCT
esrj-89750	194	1	for	for	ADP
esrj-89750	194	2	each	each	DET
esrj-89750	194	3	pixel	pixel	NOUN
esrj-89750	194	4	,	,	PUNCT
esrj-89750	194	5	it	it	PRON
esrj-89750	194	6	is	be	AUX
esrj-89750	194	7	calculated	calculate	VERB
esrj-89750	194	8	several	several	ADJ
esrj-89750	194	9	times	time	NOUN
esrj-89750	194	10	in	in	ADP
esrj-89750	194	11	several	several	ADJ
esrj-89750	194	12	categories	category	NOUN
esrj-89750	194	13	.	.	PUNCT
esrj-89750	195	1	finally	finally	ADV
esrj-89750	195	2	,	,	PUNCT
esrj-89750	195	3	the	the	DET
esrj-89750	195	4	size	size	NOUN
esrj-89750	195	5	is	be	AUX
esrj-89750	195	6	compared	compare	VERB
esrj-89750	195	7	,	,	PUNCT
esrj-89750	195	8	and	and	CCONJ
esrj-89750	195	9	the	the	DET
esrj-89750	195	10	largest	large	ADJ
esrj-89750	195	11	category	category	NOUN
esrj-89750	195	12	is	be	AUX
esrj-89750	195	13	selected	select	VERB
esrj-89750	195	14	.	.	PUNCT
esrj-89750	196	1	step	step	NOUN
esrj-89750	196	2	5	5	NUM
esrj-89750	196	3	:	:	PUNCT
esrj-89750	196	4	generate	generate	VERB
esrj-89750	196	5	a	a	DET
esrj-89750	196	6	classification	classification	NOUN
esrj-89750	196	7	map	map	NOUN
esrj-89750	196	8	and	and	CCONJ
esrj-89750	196	9	specify	specify	VERB
esrj-89750	196	10	a	a	DET
esrj-89750	196	11	value	value	NOUN
esrj-89750	196	12	for	for	ADP
esrj-89750	196	13	each	each	DET
esrj-89750	196	14	category	category	NOUN
esrj-89750	196	15	.	.	PUNCT
esrj-89750	197	1	if	if	SCONJ
esrj-89750	197	2	each	each	DET
esrj-89750	197	3	category	category	NOUN
esrj-89750	197	4	is	be	AUX
esrj-89750	197	5	divided	divide	VERB
esrj-89750	197	6	into	into	ADP
esrj-89750	197	7	10	10	NUM
esrj-89750	197	8	categories	category	NOUN
esrj-89750	197	9	,	,	PUNCT
esrj-89750	197	10	it	it	PRON
esrj-89750	197	11	is	be	AUX
esrj-89750	197	12	determined	determine	VERB
esrj-89750	197	13	that	that	SCONJ
esrj-89750	197	14	each	each	DET
esrj-89750	197	15	category	category	NOUN
esrj-89750	197	16	is	be	AUX
esrj-89750	197	17	1	1	NUM
esrj-89750	197	18	,	,	PUNCT
esrj-89750	197	19	2	2	NUM
esrj-89750	197	20	,	,	PUNCT
esrj-89750	197	21	...	...	PUNCT
esrj-89750	197	22	,	,	PUNCT
esrj-89750	197	23	10	10	NUM
esrj-89750	197	24	.	.	PUNCT
esrj-89750	198	1	the	the	DET
esrj-89750	198	2	classified	classified	ADJ
esrj-89750	198	3	pixel	pixel	PROPN
esrj-89750	198	4	values	value	NOUN
esrj-89750	198	5	are	be	AUX
esrj-89750	198	6	replaced	replace	VERB
esrj-89750	198	7	by	by	ADP
esrj-89750	198	8	category	category	NOUN
esrj-89750	198	9	values	value	NOUN
esrj-89750	198	10	.	.	PUNCT
esrj-89750	199	1	the	the	DET
esrj-89750	199	2	final	final	ADJ
esrj-89750	199	3	classified	classified	ADJ
esrj-89750	199	4	image	image	NOUN
esrj-89750	199	5	is	be	AUX
esrj-89750	199	6	thematic	thematic	ADJ
esrj-89750	199	7	image	image	NOUN
esrj-89750	199	8	.	.	PUNCT
esrj-89750	200	1	because	because	SCONJ
esrj-89750	200	2	the	the	DET
esrj-89750	200	3	maximum	maximum	ADJ
esrj-89750	200	4	gray	gray	ADJ
esrj-89750	200	5	value	value	NOUN
esrj-89750	200	6	is	be	AUX
esrj-89750	200	7	equal	equal	ADJ
esrj-89750	200	8	to	to	ADP
esrj-89750	200	9	the	the	DET
esrj-89750	200	10	number	number	NOUN
esrj-89750	200	11	of	of	ADP
esrj-89750	200	12	categories	category	NOUN
esrj-89750	200	13	,	,	PUNCT
esrj-89750	200	14	it	it	PRON
esrj-89750	200	15	needs	need	VERB
esrj-89750	200	16	to	to	PART
esrj-89750	200	17	add	add	VERB
esrj-89750	200	18	different	different	ADJ
esrj-89750	200	19	colors	color	NOUN
esrj-89750	200	20	to	to	ADP
esrj-89750	200	21	all	all	DET
esrj-89750	200	22	kinds	kind	NOUN
esrj-89750	200	23	,	,	PUNCT
esrj-89750	200	24	when	when	SCONJ
esrj-89750	200	25	displaying	display	VERB
esrj-89750	200	26	on	on	ADP
esrj-89750	200	27	the	the	DET
esrj-89750	200	28	monitor	monitor	NOUN
esrj-89750	200	29	.	.	PUNCT
esrj-89750	201	1	step	step	NOUN
esrj-89750	201	2	6	6	NUM
esrj-89750	201	3	:	:	PUNCT
esrj-89750	201	4	if	if	SCONJ
esrj-89750	201	5	there	there	PRON
esrj-89750	201	6	are	be	VERB
esrj-89750	201	7	many	many	ADJ
esrj-89750	201	8	errors	error	NOUN
esrj-89750	201	9	in	in	ADP
esrj-89750	201	10	the	the	DET
esrj-89750	201	11	classification	classification	NOUN
esrj-89750	201	12	,	,	PUNCT
esrj-89750	201	13	it	it	PRON
esrj-89750	201	14	needs	need	VERB
esrj-89750	201	15	to	to	PART
esrj-89750	201	16	re	re	VERB
esrj-89750	201	17	-	-	VERB
esrj-89750	201	18	select	select	VERB
esrj-89750	201	19	the	the	DET
esrj-89750	201	20	training	training	NOUN
esrj-89750	201	21	area	area	NOUN
esrj-89750	201	22	and	and	CCONJ
esrj-89750	201	23	do	do	VERB
esrj-89750	201	24	the	the	DET
esrj-89750	201	25	above	above	ADJ
esrj-89750	201	26	steps	step	NOUN
esrj-89750	201	27	until	until	SCONJ
esrj-89750	201	28	the	the	DET
esrj-89750	201	29	results	result	NOUN
esrj-89750	201	30	are	be	AUX
esrj-89750	201	31	satisfactory	satisfactory	ADJ
esrj-89750	201	32	.	.	PUNCT
esrj-89750	202	1	based	base	VERB
esrj-89750	202	2	on	on	ADP
esrj-89750	202	3	the	the	DET
esrj-89750	202	4	principle	principle	NOUN
esrj-89750	202	5	of	of	ADP
esrj-89750	202	6	maximum	maximum	ADJ
esrj-89750	202	7	likelihood	likelihood	NOUN
esrj-89750	202	8	method	method	NOUN
esrj-89750	202	9	,	,	PUNCT
esrj-89750	202	10	the	the	DET
esrj-89750	202	11	process	process	NOUN
esrj-89750	202	12	of	of	ADP
esrj-89750	202	13	soil	soil	NOUN
esrj-89750	202	14	remote	remote	ADJ
esrj-89750	202	15	sensing	sense	VERB
esrj-89750	202	16	image	image	NOUN
esrj-89750	202	17	classification	classification	NOUN
esrj-89750	202	18	is	be	AUX
esrj-89750	202	19	as	as	SCONJ
esrj-89750	202	20	follows	follow	VERB
esrj-89750	202	21	:	:	PUNCT
esrj-89750	202	22	step	step	NOUN
esrj-89750	202	23	1	1	NUM
esrj-89750	202	24	:	:	PUNCT
esrj-89750	202	25	preprocess	preprocess	VERB
esrj-89750	202	26	the	the	DET
esrj-89750	202	27	image	image	NOUN
esrj-89750	202	28	;	;	PUNCT
esrj-89750	202	29	step	step	NOUN
esrj-89750	202	30	2	2	NUM
esrj-89750	202	31	:	:	PUNCT
esrj-89750	202	32	initialize	initialize	VERB
esrj-89750	202	33	the	the	DET
esrj-89750	202	34	number	number	NOUN
esrj-89750	202	35	of	of	ADP
esrj-89750	202	36	classifications	classification	NOUN
esrj-89750	202	37	and	and	CCONJ
esrj-89750	202	38	training	training	NOUN
esrj-89750	202	39	parameters	parameter	NOUN
esrj-89750	202	40	to	to	PART
esrj-89750	202	41	determine	determine	VERB
esrj-89750	202	42	the	the	DET
esrj-89750	202	43	number	number	NOUN
esrj-89750	202	44	of	of	ADP
esrj-89750	202	45	data	datum	NOUN
esrj-89750	202	46	blocks	block	NOUN
esrj-89750	202	47	and	and	CCONJ
esrj-89750	202	48	threads	thread	NOUN
esrj-89750	202	49	;	;	PUNCT
esrj-89750	202	50	step	step	NOUN
esrj-89750	202	51	3	3	NUM
esrj-89750	202	52	:	:	PUNCT
esrj-89750	202	53	establish	establish	VERB
esrj-89750	202	54	and	and	CCONJ
esrj-89750	202	55	initialize	initialize	VERB
esrj-89750	202	56	grid	grid	NOUN
esrj-89750	202	57	threads	thread	NOUN
esrj-89750	202	58	and	and	CCONJ
esrj-89750	202	59	the	the	DET
esrj-89750	202	60	amount	amount	NOUN
esrj-89750	202	61	of	of	ADP
esrj-89750	202	62	data	datum	NOUN
esrj-89750	202	63	to	to	PART
esrj-89750	202	64	be	be	AUX
esrj-89750	202	65	computed	compute	VERB
esrj-89750	202	66	by	by	ADP
esrj-89750	202	67	each	each	DET
esrj-89750	202	68	thread	thread	NOUN
esrj-89750	202	69	.	.	PUNCT
esrj-89750	203	1	step	step	NOUN
esrj-89750	203	2	4	4	NUM
esrj-89750	203	3	:	:	PUNCT
esrj-89750	203	4	detect	detect	VERB
esrj-89750	203	5	the	the	DET
esrj-89750	203	6	connectivity	connectivity	NOUN
esrj-89750	203	7	of	of	ADP
esrj-89750	203	8	the	the	DET
esrj-89750	203	9	network	network	NOUN
esrj-89750	203	10	and	and	CCONJ
esrj-89750	203	11	grid	grid	NOUN
esrj-89750	203	12	nodes	node	NOUN
esrj-89750	203	13	,	,	PUNCT
esrj-89750	203	14	and	and	CCONJ
esrj-89750	203	15	add	add	VERB
esrj-89750	203	16	a	a	DET
esrj-89750	203	17	grid	grid	NOUN
esrj-89750	203	18	computing	computing	NOUN
esrj-89750	203	19	thread	thread	NOUN
esrj-89750	203	20	model	model	NOUN
esrj-89750	203	21	to	to	ADP
esrj-89750	203	22	the	the	DET
esrj-89750	203	23	grid	grid	NOUN
esrj-89750	203	24	;	;	PUNCT
esrj-89750	203	25	step	step	NOUN
esrj-89750	203	26	5	5	NUM
esrj-89750	203	27	:	:	PUNCT
esrj-89750	203	28	each	each	DET
esrj-89750	203	29	thread	thread	NOUN
esrj-89750	203	30	carries	carry	VERB
esrj-89750	203	31	out	out	ADP
esrj-89750	203	32	cyclic	cyclic	ADJ
esrj-89750	203	33	grid	grid	NOUN
esrj-89750	203	34	calculation	calculation	NOUN
esrj-89750	203	35	according	accord	VERB
esrj-89750	203	36	to	to	ADP
esrj-89750	203	37	the	the	DET
esrj-89750	203	38	sample	sample	NOUN
esrj-89750	203	39	’s	’s	PART
esrj-89750	203	40	subset	subset	ADJ
esrj-89750	203	41	data	datum	NOUN
esrj-89750	203	42	,	,	PUNCT
esrj-89750	203	43	and	and	CCONJ
esrj-89750	203	44	all	all	DET
esrj-89750	203	45	data	datum	NOUN
esrj-89750	203	46	in	in	ADP
esrj-89750	203	47	the	the	DET
esrj-89750	203	48	grid	grid	NOUN
esrj-89750	203	49	unit	unit	NOUN
esrj-89750	203	50	are	be	AUX
esrj-89750	203	51	computed	compute	VERB
esrj-89750	203	52	to	to	PART
esrj-89750	203	53	get	get	VERB
esrj-89750	203	54	the	the	DET
esrj-89750	203	55	category	category	NOUN
esrj-89750	203	56	and	and	CCONJ
esrj-89750	203	57	stored	store	VERB
esrj-89750	203	58	to	to	ADP
esrj-89750	203	59	this	this	DET
esrj-89750	203	60	node	node	NOUN
esrj-89750	203	61	.	.	PUNCT
esrj-89750	204	1	step	step	NOUN
esrj-89750	204	2	6	6	NUM
esrj-89750	204	3	:	:	PUNCT
esrj-89750	204	4	transfer	transfer	VERB
esrj-89750	204	5	the	the	DET
esrj-89750	204	6	results	result	NOUN
esrj-89750	204	7	from	from	ADP
esrj-89750	204	8	each	each	DET
esrj-89750	204	9	grid	grid	NOUN
esrj-89750	204	10	site	site	NOUN
esrj-89750	204	11	to	to	ADP
esrj-89750	204	12	the	the	DET
esrj-89750	204	13	host	host	NOUN
esrj-89750	204	14	computer	computer	NOUN
esrj-89750	204	15	,	,	PUNCT
esrj-89750	204	16	and	and	CCONJ
esrj-89750	204	17	merge	merge	VERB
esrj-89750	204	18	the	the	DET
esrj-89750	204	19	grid	grid	NOUN
esrj-89750	204	20	results	result	NOUN
esrj-89750	204	21	.	.	PUNCT
esrj-89750	205	1	step	step	NOUN
esrj-89750	205	2	7	7	NUM
esrj-89750	205	3	:	:	PUNCT
esrj-89750	205	4	release	release	VERB
esrj-89750	205	5	grid	grid	NOUN
esrj-89750	205	6	resources	resource	NOUN
esrj-89750	205	7	and	and	CCONJ
esrj-89750	205	8	output	output	NOUN
esrj-89750	205	9	classification	classification	NOUN
esrj-89750	205	10	results	result	NOUN
esrj-89750	205	11	to	to	ADP
esrj-89750	205	12	files	file	NOUN
esrj-89750	205	13	.	.	PUNCT
esrj-89750	206	1	example	example	NOUN
esrj-89750	206	2	analysis	analysis	NOUN
esrj-89750	206	3	general	general	ADJ
esrj-89750	206	4	situation	situation	NOUN
esrj-89750	206	5	of	of	ADP
esrj-89750	206	6	test	test	NOUN
esrj-89750	206	7	area	area	NOUN
esrj-89750	206	8	lindian	lindian	PROPN
esrj-89750	206	9	county	county	PROPN
esrj-89750	206	10	is	be	AUX
esrj-89750	206	11	located	locate	VERB
esrj-89750	206	12	in	in	ADP
esrj-89750	206	13	the	the	DET
esrj-89750	206	14	western	western	ADJ
esrj-89750	206	15	part	part	NOUN
esrj-89750	206	16	of	of	ADP
esrj-89750	206	17	heilongjiang	heilongjiang	PROPN
esrj-89750	206	18	province	province	PROPN
esrj-89750	206	19	,	,	PUNCT
esrj-89750	206	20	the	the	DET
esrj-89750	206	21	hinterland	hinterland	NOUN
esrj-89750	206	22	of	of	ADP
esrj-89750	206	23	songnen	songnen	NOUN
esrj-89750	206	24	plain	plain	ADJ
esrj-89750	206	25	,	,	PUNCT
esrj-89750	206	26	with	with	ADP
esrj-89750	206	27	an	an	DET
esrj-89750	206	28	area	area	NOUN
esrj-89750	206	29	of	of	ADP
esrj-89750	206	30	about	about	ADV
esrj-89750	206	31	3	3	NUM
esrj-89750	206	32	500	500	NUM
esrj-89750	206	33	km2	km2	NOUN
esrj-89750	206	34	.	.	PUNCT
esrj-89750	207	1	it	it	PRON
esrj-89750	207	2	belongs	belong	VERB
esrj-89750	207	3	to	to	ADP
esrj-89750	207	4	daqing	daqe	VERB
esrj-89750	207	5	city	city	NOUN
esrj-89750	207	6	,	,	PUNCT
esrj-89750	207	7	heilongjiang	heilongjiang	PROPN
esrj-89750	207	8	province	province	NOUN
esrj-89750	207	9	.	.	PUNCT
esrj-89750	208	1	it	it	PRON
esrj-89750	208	2	has	have	VERB
esrj-89750	208	3	a	a	DET
esrj-89750	208	4	continental	continental	ADJ
esrj-89750	208	5	monsoon	monsoon	NOUN
esrj-89750	208	6	climate	climate	NOUN
esrj-89750	208	7	in	in	ADP
esrj-89750	208	8	the	the	DET
esrj-89750	208	9	middle	middle	ADJ
esrj-89750	208	10	temperate	temperate	ADJ
esrj-89750	208	11	zone	zone	NOUN
esrj-89750	208	12	.	.	PUNCT
esrj-89750	209	1	it	it	PRON
esrj-89750	209	2	is	be	AUX
esrj-89750	209	3	rainy	rainy	ADJ
esrj-89750	209	4	and	and	CCONJ
esrj-89750	209	5	warm	warm	ADJ
esrj-89750	209	6	in	in	ADP
esrj-89750	209	7	summer	summer	NOUN
esrj-89750	209	8	,	,	PUNCT
esrj-89750	209	9	dry	dry	ADJ
esrj-89750	209	10	and	and	CCONJ
esrj-89750	209	11	cold	cold	ADJ
esrj-89750	209	12	in	in	ADP
esrj-89750	209	13	winter	winter	NOUN
esrj-89750	209	14	.	.	PUNCT
esrj-89750	210	1	its	its	PRON
esrj-89750	210	2	geological	geological	ADJ
esrj-89750	210	3	structure	structure	NOUN
esrj-89750	210	4	belongs	belong	VERB
esrj-89750	210	5	to	to	ADP
esrj-89750	210	6	the	the	DET
esrj-89750	210	7	paleo	paleo	ADJ
esrj-89750	210	8	-	-	PUNCT
esrj-89750	210	9	asian	asian	ADJ
esrj-89750	210	10	tectonic	tectonic	ADJ
esrj-89750	210	11	domain	domain	NOUN
esrj-89750	210	12	.	.	PUNCT
esrj-89750	211	1	it	it	PRON
esrj-89750	211	2	has	have	VERB
esrj-89750	211	3	abundant	abundant	ADJ
esrj-89750	211	4	wetland	wetland	NOUN
esrj-89750	211	5	resources	resource	NOUN
esrj-89750	211	6	and	and	CCONJ
esrj-89750	211	7	natural	natural	ADJ
esrj-89750	211	8	rivers	river	NOUN
esrj-89750	211	9	such	such	ADJ
esrj-89750	211	10	as	as	ADP
esrj-89750	211	11	uyur	uyur	PROPN
esrj-89750	211	12	river	river	PROPN
esrj-89750	211	13	and	and	CCONJ
esrj-89750	211	14	shuangyang	shuangyang	PROPN
esrj-89750	211	15	river	river	PROPN
esrj-89750	211	16	.	.	PUNCT
esrj-89750	212	1	the	the	DET
esrj-89750	212	2	elevation	elevation	NOUN
esrj-89750	212	3	is	be	AUX
esrj-89750	212	4	low	low	ADJ
esrj-89750	212	5	and	and	CCONJ
esrj-89750	212	6	terrain	terrain	NOUN
esrj-89750	212	7	is	be	AUX
esrj-89750	212	8	flat	flat	ADJ
esrj-89750	212	9	,	,	PUNCT
esrj-89750	212	10	but	but	CCONJ
esrj-89750	212	11	microtopography	microtopography	NOUN
esrj-89750	212	12	is	be	AUX
esrj-89750	212	13	more	more	ADV
esrj-89750	212	14	complex	complex	ADJ
esrj-89750	212	15	and	and	CCONJ
esrj-89750	212	16	low	low	ADV
esrj-89750	212	17	-	-	PUNCT
esrj-89750	212	18	lying	lie	VERB
esrj-89750	212	19	leads	lead	NOUN
esrj-89750	212	20	to	to	ADP
esrj-89750	212	21	difficult	difficult	ADJ
esrj-89750	212	22	drainage	drainage	NOUN
esrj-89750	212	23	,	,	PUNCT
esrj-89750	212	24	which	which	PRON
esrj-89750	212	25	is	be	AUX
esrj-89750	212	26	easy	easy	ADJ
esrj-89750	212	27	to	to	PART
esrj-89750	212	28	form	form	VERB
esrj-89750	212	29	marshes	marsh	NOUN
esrj-89750	212	30	.	.	PUNCT
esrj-89750	213	1	according	accord	VERB
esrj-89750	213	2	to	to	ADP
esrj-89750	213	3	the	the	DET
esrj-89750	213	4	data	datum	NOUN
esrj-89750	213	5	of	of	ADP
esrj-89750	213	6	the	the	DET
esrj-89750	213	7	second	second	ADJ
esrj-89750	213	8	national	national	ADJ
esrj-89750	213	9	soil	soil	NOUN
esrj-89750	213	10	census	census	NOUN
esrj-89750	213	11	,	,	PUNCT
esrj-89750	213	12	most	most	ADJ
esrj-89750	213	13	of	of	ADP
esrj-89750	213	14	the	the	DET
esrj-89750	213	15	parent	parent	NOUN
esrj-89750	213	16	materials	material	NOUN
esrj-89750	213	17	of	of	ADP
esrj-89750	213	18	the	the	DET
esrj-89750	213	19	soils	soil	NOUN
esrj-89750	213	20	in	in	ADP
esrj-89750	213	21	lindian	lindian	PROPN
esrj-89750	213	22	county	county	NOUN
esrj-89750	213	23	are	be	AUX
esrj-89750	213	24	loesslike	loesslike	ADJ
esrj-89750	213	25	deposits	deposit	NOUN
esrj-89750	213	26	(	(	PUNCT
esrj-89750	213	27	palagan	palagan	NOUN
esrj-89750	213	28	&	&	CCONJ
esrj-89750	213	29	geetha	geetha	PROPN
esrj-89750	213	30	,	,	PUNCT
esrj-89750	213	31	2016	2016	NUM
esrj-89750	213	32	)	)	PUNCT
esrj-89750	213	33	.	.	PUNCT
esrj-89750	214	1	there	there	PRON
esrj-89750	214	2	are	be	VERB
esrj-89750	214	3	four	four	NUM
esrj-89750	214	4	main	main	ADJ
esrj-89750	214	5	soil	soil	NOUN
esrj-89750	214	6	types	type	NOUN
esrj-89750	214	7	(	(	PUNCT
esrj-89750	214	8	by	by	ADP
esrj-89750	214	9	the	the	DET
esrj-89750	214	10	second	second	ADJ
esrj-89750	214	11	national	national	ADJ
esrj-89750	214	12	soil	soil	NOUN
esrj-89750	214	13	census	census	NOUN
esrj-89750	214	14	)	)	PUNCT
esrj-89750	214	15	.	.	PUNCT
esrj-89750	215	1	they	they	PRON
esrj-89750	215	2	are	be	AUX
esrj-89750	215	3	chernozem	chernozem	ADJ
esrj-89750	215	4	,	,	PUNCT
esrj-89750	215	5	meadow	meadow	NOUN
esrj-89750	215	6	soil	soil	NOUN
esrj-89750	215	7	,	,	PUNCT
esrj-89750	215	8	marsh	marsh	ADJ
esrj-89750	215	9	soil	soil	NOUN
esrj-89750	215	10	and	and	CCONJ
esrj-89750	215	11	aeolian	aeolian	ADJ
esrj-89750	215	12	sandy	sandy	ADJ
esrj-89750	215	13	soil	soil	NOUN
esrj-89750	215	14	.	.	PUNCT
esrj-89750	216	1	among	among	ADP
esrj-89750	216	2	them	they	PRON
esrj-89750	216	3	,	,	PUNCT
esrj-89750	216	4	chernozem	chernozem	NOUN
esrj-89750	216	5	accounts	account	VERB
esrj-89750	216	6	for	for	ADP
esrj-89750	216	7	more	more	ADJ
esrj-89750	216	8	than	than	ADP
esrj-89750	216	9	60	60	NUM
esrj-89750	216	10	%	%	NOUN
esrj-89750	216	11	of	of	ADP
esrj-89750	216	12	the	the	DET
esrj-89750	216	13	total	total	ADJ
esrj-89750	216	14	soil	soil	NOUN
esrj-89750	216	15	.	.	PUNCT
esrj-89750	217	1	chernozem	chernozem	PROPN
esrj-89750	217	2	is	be	AUX
esrj-89750	217	3	also	also	ADV
esrj-89750	217	4	an	an	DET
esrj-89750	217	5	important	important	ADJ
esrj-89750	217	6	part	part	NOUN
esrj-89750	217	7	of	of	ADP
esrj-89750	217	8	black	black	ADJ
esrj-89750	217	9	soil	soil	NOUN
esrj-89750	217	10	resources	resource	NOUN
esrj-89750	217	11	,	,	PUNCT
esrj-89750	217	12	and	and	CCONJ
esrj-89750	217	13	its	its	PRON
esrj-89750	217	14	organic	organic	ADJ
esrj-89750	217	15	matter	matter	NOUN
esrj-89750	217	16	content	content	NOUN
esrj-89750	217	17	is	be	AUX
esrj-89750	217	18	high	high	ADJ
esrj-89750	217	19	.	.	PUNCT
esrj-89750	218	1	soil	soil	NOUN
esrj-89750	218	2	fertility	fertility	NOUN
esrj-89750	218	3	is	be	AUX
esrj-89750	218	4	large	large	ADJ
esrj-89750	218	5	,	,	PUNCT
esrj-89750	218	6	which	which	PRON
esrj-89750	218	7	is	be	AUX
esrj-89750	218	8	a	a	DET
esrj-89750	218	9	very	very	ADV
esrj-89750	218	10	suitable	suitable	ADJ
esrj-89750	218	11	soil	soil	NOUN
esrj-89750	218	12	type	type	NOUN
esrj-89750	218	13	for	for	ADP
esrj-89750	218	14	grain	grain	NOUN
esrj-89750	218	15	production	production	NOUN
esrj-89750	218	16	.	.	PUNCT
esrj-89750	219	1	the	the	DET
esrj-89750	219	2	landsat	landsat	PROPN
esrj-89750	219	3	8	8	NUM
esrj-89750	219	4	oli	oli	ADJ
esrj-89750	219	5	image	image	NOUN
esrj-89750	219	6	of	of	ADP
esrj-89750	219	7	lindian	lindian	ADJ
esrj-89750	219	8	area	area	NOUN
esrj-89750	219	9	on	on	ADP
esrj-89750	219	10	may	may	PROPN
esrj-89750	219	11	3	3	NUM
esrj-89750	219	12	,	,	PUNCT
esrj-89750	219	13	2014	2014	NUM
esrj-89750	219	14	is	be	AUX
esrj-89750	219	15	selected	select	VERB
esrj-89750	219	16	for	for	ADP
esrj-89750	219	17	this	this	DET
esrj-89750	219	18	test	test	NOUN
esrj-89750	219	19	.	.	PUNCT
esrj-89750	220	1	the	the	DET
esrj-89750	220	2	study	study	NOUN
esrj-89750	220	3	area	area	NOUN
esrj-89750	220	4	in	in	ADP
esrj-89750	220	5	the	the	DET
esrj-89750	220	6	image	image	NOUN
esrj-89750	220	7	is	be	AUX
esrj-89750	220	8	almost	almost	ADV
esrj-89750	220	9	covered	cover	VERB
esrj-89750	220	10	by	by	ADP
esrj-89750	220	11	clouds	cloud	NOUN
esrj-89750	220	12	.	.	PUNCT
esrj-89750	221	1	in	in	ADP
esrj-89750	221	2	may	may	PROPN
esrj-89750	221	3	,	,	PUNCT
esrj-89750	221	4	it	it	PRON
esrj-89750	221	5	is	be	AUX
esrj-89750	221	6	bare	bare	ADJ
esrj-89750	221	7	soil	soil	NOUN
esrj-89750	221	8	period	period	NOUN
esrj-89750	221	9	.	.	PUNCT
esrj-89750	222	1	there	there	PRON
esrj-89750	222	2	is	be	VERB
esrj-89750	222	3	neither	neither	CCONJ
esrj-89750	222	4	a	a	DET
esrj-89750	222	5	large	large	ADJ
esrj-89750	222	6	area	area	NOUN
esrj-89750	222	7	of	of	ADP
esrj-89750	222	8	vegetation	vegetation	NOUN
esrj-89750	222	9	nor	nor	CCONJ
esrj-89750	222	10	snow	snow	NOUN
esrj-89750	222	11	.	.	PUNCT
esrj-89750	223	1	it	it	PRON
esrj-89750	223	2	meets	meet	VERB
esrj-89750	223	3	the	the	DET
esrj-89750	223	4	research	research	NOUN
esrj-89750	223	5	requirements	requirement	NOUN
esrj-89750	223	6	of	of	ADP
esrj-89750	223	7	this	this	DET
esrj-89750	223	8	paper	paper	NOUN
esrj-89750	223	9	for	for	ADP
esrj-89750	223	10	bare	bare	ADJ
esrj-89750	223	11	soil	soil	NOUN
esrj-89750	223	12	period	period	NOUN
esrj-89750	223	13	,	,	PUNCT
esrj-89750	223	14	as	as	SCONJ
esrj-89750	223	15	shown	show	VERB
esrj-89750	223	16	in	in	ADP
esrj-89750	223	17	figure	figure	NOUN
esrj-89750	223	18	11	11	NUM
esrj-89750	223	19	.	.	PUNCT
esrj-89750	224	1	figure	figure	VERB
esrj-89750	224	2	11	11	NUM
esrj-89750	224	3	.	.	PUNCT
esrj-89750	225	1	test	test	NOUN
esrj-89750	225	2	zone	zone	NOUN
esrj-89750	225	3	experimental	experimental	ADJ
esrj-89750	225	4	environment	environment	NOUN
esrj-89750	225	5	and	and	CCONJ
esrj-89750	225	6	methods	method	NOUN
esrj-89750	225	7	in	in	ADP
esrj-89750	225	8	this	this	DET
esrj-89750	225	9	paper	paper	NOUN
esrj-89750	225	10	,	,	PUNCT
esrj-89750	225	11	tensor	tensor	NOUN
esrj-89750	225	12	flow	flow	NOUN
esrj-89750	225	13	machine	machine	NOUN
esrj-89750	225	14	learning	learn	VERB
esrj-89750	225	15	framework	framework	NOUN
esrj-89750	225	16	under	under	ADP
esrj-89750	225	17	linux	linux	PROPN
esrj-89750	225	18	is	be	AUX
esrj-89750	225	19	used	use	VERB
esrj-89750	225	20	,	,	PUNCT
esrj-89750	225	21	and	and	CCONJ
esrj-89750	225	22	python	python	NOUN
esrj-89750	225	23	language	language	NOUN
esrj-89750	225	24	is	be	AUX
esrj-89750	225	25	used	use	VERB
esrj-89750	225	26	for	for	ADP
esrj-89750	225	27	programming	programming	NOUN
esrj-89750	225	28	.	.	PUNCT
esrj-89750	226	1	the	the	DET
esrj-89750	226	2	hardware	hardware	NOUN
esrj-89750	226	3	environment	environment	NOUN
esrj-89750	226	4	is	be	AUX
esrj-89750	226	5	intel	intel	NOUN
esrj-89750	226	6	(	(	PUNCT
esrj-89750	226	7	r	r	NOUN
esrj-89750	226	8	)	)	PUNCT
esrj-89750	226	9	xeon	xeon	NOUN
esrj-89750	227	1	(	(	PUNCT
esrj-89750	227	2	r	r	NOUN
esrj-89750	227	3	)	)	PUNCT
esrj-89750	227	4	e5	e5	NOUN
esrj-89750	227	5	-	-	PUNCT
esrj-89750	227	6	2630	2630	NUM
esrj-89750	227	7	cpu	cpu	NOUN
esrj-89750	227	8	,	,	PUNCT
esrj-89750	227	9	nvidia	nvidia	PROPN
esrj-89750	227	10	tesla	tesla	PROPN
esrj-89750	227	11	m40	m40	PROPN
esrj-89750	227	12	upu	upu	PROPN
esrj-89750	227	13	and	and	CCONJ
esrj-89750	227	14	12	12	NUM
esrj-89750	227	15	gb	gb	NOUN
esrj-89750	227	16	memory	memory	NOUN
esrj-89750	227	17	.	.	PUNCT
esrj-89750	228	1	in	in	ADP
esrj-89750	228	2	order	order	NOUN
esrj-89750	228	3	to	to	PART
esrj-89750	228	4	verify	verify	VERB
esrj-89750	228	5	the	the	DET
esrj-89750	228	6	effectiveness	effectiveness	NOUN
esrj-89750	228	7	of	of	ADP
esrj-89750	228	8	the	the	DET
esrj-89750	228	9	proposed	propose	VERB
esrj-89750	228	10	method	method	NOUN
esrj-89750	228	11	,	,	PUNCT
esrj-89750	228	12	three	three	NUM
esrj-89750	228	13	groups	group	NOUN
esrj-89750	228	14	of	of	ADP
esrj-89750	228	15	experiments	experiment	NOUN
esrj-89750	228	16	are	be	AUX
esrj-89750	228	17	set	set	VERB
esrj-89750	228	18	up	up	ADP
esrj-89750	228	19	.	.	PUNCT
esrj-89750	229	1	the	the	DET
esrj-89750	229	2	first	first	ADJ
esrj-89750	229	3	group	group	NOUN
esrj-89750	229	4	of	of	ADP
esrj-89750	229	5	experimental	experimental	ADJ
esrj-89750	229	6	data	datum	NOUN
esrj-89750	229	7	uses	use	VERB
esrj-89750	229	8	the	the	DET
esrj-89750	229	9	proposed	propose	VERB
esrj-89750	229	10	method	method	NOUN
esrj-89750	229	11	to	to	PART
esrj-89750	229	12	classify	classify	VERB
esrj-89750	229	13	soil	soil	NOUN
esrj-89750	229	14	remote	remote	ADJ
esrj-89750	229	15	sensing	sensing	NOUN
esrj-89750	229	16	images	image	NOUN
esrj-89750	229	17	,	,	PUNCT
esrj-89750	229	18	the	the	DET
esrj-89750	229	19	second	second	ADJ
esrj-89750	229	20	group	group	NOUN
esrj-89750	229	21	uses	use	VERB
esrj-89750	229	22	deep	deep	ADJ
esrj-89750	229	23	learning	learning	NOUN
esrj-89750	229	24	-	-	PUNCT
esrj-89750	229	25	based	base	VERB
esrj-89750	229	26	methods	method	NOUN
esrj-89750	229	27	to	to	PART
esrj-89750	229	28	classify	classify	VERB
esrj-89750	229	29	soil	soil	NOUN
esrj-89750	229	30	remote	remote	ADJ
esrj-89750	229	31	sensing	sensing	NOUN
esrj-89750	229	32	images	image	NOUN
esrj-89750	229	33	,	,	PUNCT
esrj-89750	229	34	and	and	CCONJ
esrj-89750	229	35	the	the	DET
esrj-89750	229	36	third	third	ADJ
esrj-89750	229	37	group	group	NOUN
esrj-89750	229	38	uses	use	VERB
esrj-89750	229	39	maximum	maximum	ADJ
esrj-89750	229	40	likelihood	likelihood	NOUN
esrj-89750	229	41	estimation	estimation	NOUN
esrj-89750	229	42	to	to	PART
esrj-89750	229	43	classify	classify	VERB
esrj-89750	229	44	soil	soil	NOUN
esrj-89750	229	45	remote	remote	ADJ
esrj-89750	229	46	sensing	sensing	NOUN
esrj-89750	229	47	images	image	NOUN
esrj-89750	229	48	.	.	PUNCT
esrj-89750	230	1	sample	sample	NOUN
esrj-89750	230	2	training	training	NOUN
esrj-89750	230	3	samples	sample	NOUN
esrj-89750	230	4	are	be	AUX
esrj-89750	230	5	divided	divide	VERB
esrj-89750	230	6	into	into	ADP
esrj-89750	230	7	four	four	NUM
esrj-89750	230	8	categories	category	NOUN
esrj-89750	230	9	:	:	PUNCT
esrj-89750	230	10	chernozem	chernozem	ADJ
esrj-89750	230	11	,	,	PUNCT
esrj-89750	230	12	meadow	meadow	NOUN
esrj-89750	230	13	soil	soil	NOUN
esrj-89750	230	14	,	,	PUNCT
esrj-89750	230	15	marsh	marsh	ADJ
esrj-89750	230	16	soil	soil	NOUN
esrj-89750	230	17	and	and	CCONJ
esrj-89750	230	18	aeolian	aeolian	ADJ
esrj-89750	230	19	sandy	sandy	ADJ
esrj-89750	230	20	soil	soil	NOUN
esrj-89750	230	21	.	.	PUNCT
esrj-89750	231	1	there	there	PRON
esrj-89750	231	2	are	be	VERB
esrj-89750	231	3	10	10	NUM
esrj-89750	231	4	samples	sample	NOUN
esrj-89750	231	5	in	in	ADP
esrj-89750	231	6	each	each	DET
esrj-89750	231	7	category	category	NOUN
esrj-89750	231	8	and	and	CCONJ
esrj-89750	231	9	80	80	NUM
esrj-89750	231	10	samples	sample	NOUN
esrj-89750	231	11	in	in	ADP
esrj-89750	231	12	total	total	NOUN
esrj-89750	231	13	.	.	PUNCT
esrj-89750	232	1	40	40	NUM
esrj-89750	232	2	samples	sample	NOUN
esrj-89750	232	3	are	be	AUX
esrj-89750	232	4	used	use	VERB
esrj-89750	232	5	for	for	ADP
esrj-89750	232	6	experimental	experimental	ADJ
esrj-89750	232	7	training	training	NOUN
esrj-89750	232	8	and	and	CCONJ
esrj-89750	232	9	the	the	DET
esrj-89750	232	10	remaining	remain	VERB
esrj-89750	232	11	40	40	NUM
esrj-89750	232	12	samples	sample	NOUN
esrj-89750	232	13	are	be	AUX
esrj-89750	232	14	used	use	VERB
esrj-89750	232	15	for	for	ADP
esrj-89750	232	16	testing	testing	NOUN
esrj-89750	232	17	.	.	PUNCT
esrj-89750	233	1	the	the	DET
esrj-89750	233	2	parameters	parameter	NOUN
esrj-89750	233	3	of	of	ADP
esrj-89750	233	4	the	the	DET
esrj-89750	233	5	trained	train	VERB
esrj-89750	233	6	convolutional	convolutional	ADJ
esrj-89750	233	7	neural	neural	ADJ
esrj-89750	233	8	network	network	NOUN
esrj-89750	233	9	are	be	AUX
esrj-89750	233	10	shown	show	VERB
esrj-89750	233	11	in	in	ADP
esrj-89750	233	12	table	table	NOUN
esrj-89750	233	13	3	3	NUM
esrj-89750	233	14	.	.	PUNCT
esrj-89750	233	15	test	test	NOUN
esrj-89750	233	16	verification	verification	NOUN
esrj-89750	233	17	in	in	ADP
esrj-89750	233	18	the	the	DET
esrj-89750	233	19	trained	train	VERB
esrj-89750	233	20	model	model	NOUN
esrj-89750	233	21	,	,	PUNCT
esrj-89750	233	22	the	the	DET
esrj-89750	233	23	remaining	remain	VERB
esrj-89750	233	24	sample	sample	NOUN
esrj-89750	233	25	data	data	NOUN
esrj-89750	233	26	sets	set	NOUN
esrj-89750	233	27	are	be	AUX
esrj-89750	233	28	input	input	NOUN
esrj-89750	233	29	for	for	ADP
esrj-89750	233	30	a	a	DET
esrj-89750	233	31	performance	performance	NOUN
esrj-89750	233	32	test	test	NOUN
esrj-89750	233	33	,	,	PUNCT
esrj-89750	233	34	and	and	CCONJ
esrj-89750	233	35	the	the	DET
esrj-89750	233	36	experimental	experimental	ADJ
esrj-89750	233	37	results	result	NOUN
esrj-89750	233	38	are	be	AUX
esrj-89750	233	39	expressed	express	VERB
esrj-89750	233	40	in	in	ADP
esrj-89750	233	41	the	the	DET
esrj-89750	233	42	form	form	NOUN
esrj-89750	233	43	of	of	ADP
esrj-89750	233	44	confusion	confusion	NOUN
esrj-89750	233	45	matrix	matrix	NOUN
esrj-89750	233	46	.	.	PUNCT
esrj-89750	234	1	confusion	confusion	NOUN
esrj-89750	234	2	matrix	matrix	NOUN
esrj-89750	234	3	,	,	PUNCT
esrj-89750	234	4	also	also	ADV
esrj-89750	234	5	known	know	VERB
esrj-89750	234	6	as	as	ADP
esrj-89750	234	7	error	error	NOUN
esrj-89750	234	8	matrix	matrix	NOUN
esrj-89750	234	9	,	,	PUNCT
esrj-89750	234	10	is	be	AUX
esrj-89750	234	11	a	a	DET
esrj-89750	234	12	standard	standard	ADJ
esrj-89750	234	13	format	format	NOUN
esrj-89750	234	14	for	for	ADP
esrj-89750	234	15	accuracy	accuracy	NOUN
esrj-89750	234	16	evaluation	evaluation	NOUN
esrj-89750	234	17	,	,	PUNCT
esrj-89750	234	18	which	which	PRON
esrj-89750	234	19	is	be	AUX
esrj-89750	234	20	expressed	express	VERB
esrj-89750	234	21	in	in	ADP
esrj-89750	234	22	the	the	DET
esrj-89750	234	23	matrix	matrix	NOUN
esrj-89750	234	24	form	form	NOUN
esrj-89750	234	25	of	of	ADP
esrj-89750	234	26	n	n	NOUN
esrj-89750	234	27	rows	row	NOUN
esrj-89750	234	28	and	and	CCONJ
esrj-89750	234	29	n	n	DET
esrj-89750	234	30	columns	column	NOUN
esrj-89750	234	31	.	.	PUNCT
esrj-89750	235	1	specific	specific	ADJ
esrj-89750	235	2	evaluation	evaluation	NOUN
esrj-89750	235	3	indicators	indicator	NOUN
esrj-89750	235	4	include	include	VERB
esrj-89750	235	5	overall	overall	ADJ
esrj-89750	235	6	accuracy	accuracy	NOUN
esrj-89750	235	7	,	,	PUNCT
esrj-89750	235	8	mapping	mapping	NOUN
esrj-89750	235	9	accuracy	accuracy	NOUN
esrj-89750	235	10	,	,	PUNCT
esrj-89750	235	11	user	user	NOUN
esrj-89750	235	12	accuracy	accuracy	NOUN
esrj-89750	235	13	and	and	CCONJ
esrj-89750	235	14	so	so	ADV
esrj-89750	235	15	on	on	ADV
esrj-89750	235	16	.	.	PUNCT
esrj-89750	236	1	these	these	DET
esrj-89750	236	2	accuracy	accuracy	NOUN
esrj-89750	236	3	indicators	indicator	NOUN
esrj-89750	236	4	reflect	reflect	VERB
esrj-89750	236	5	the	the	DET
esrj-89750	236	6	accuracy	accuracy	NOUN
esrj-89750	236	7	of	of	ADP
esrj-89750	236	8	image	image	NOUN
esrj-89750	236	9	classification	classification	NOUN
esrj-89750	236	10	from	from	ADP
esrj-89750	236	11	different	different	ADJ
esrj-89750	236	12	aspects	aspect	NOUN
esrj-89750	236	13	(	(	PUNCT
esrj-89750	236	14	sato	sato	NOUN
esrj-89750	236	15	,	,	PUNCT
esrj-89750	236	16	2012	2012	NUM
esrj-89750	236	17	;	;	PUNCT
esrj-89750	236	18	zhu	zhu	PROPN
esrj-89750	236	19	,	,	PUNCT
esrj-89750	236	20	2016	2016	NUM
esrj-89750	236	21	)	)	PUNCT
esrj-89750	236	22	.	.	PUNCT
esrj-89750	237	1	each	each	DET
esrj-89750	237	2	column	column	NOUN
esrj-89750	237	3	of	of	ADP
esrj-89750	237	4	confusion	confusion	NOUN
esrj-89750	237	5	matrix	matrix	NOUN
esrj-89750	237	6	represents	represent	VERB
esrj-89750	237	7	the	the	DET
esrj-89750	237	8	prediction	prediction	NOUN
esrj-89750	237	9	category	category	NOUN
esrj-89750	237	10	,	,	PUNCT
esrj-89750	237	11	and	and	CCONJ
esrj-89750	237	12	the	the	DET
esrj-89750	237	13	total	total	ADJ
esrj-89750	237	14	number	number	NOUN
esrj-89750	237	15	of	of	ADP
esrj-89750	237	16	each	each	DET
esrj-89750	237	17	column	column	NOUN
esrj-89750	237	18	represents	represent	VERB
esrj-89750	237	19	the	the	DET
esrj-89750	237	20	number	number	NOUN
esrj-89750	237	21	of	of	ADP
esrj-89750	237	22	data	datum	NOUN
esrj-89750	237	23	predicted	predict	VERB
esrj-89750	237	24	for	for	ADP
esrj-89750	237	25	that	that	DET
esrj-89750	237	26	category	category	NOUN
esrj-89750	237	27	;	;	PUNCT
esrj-89750	237	28	each	each	DET
esrj-89750	237	29	row	row	NOUN
esrj-89750	237	30	represents	represent	VERB
esrj-89750	237	31	the	the	DET
esrj-89750	237	32	true	true	ADJ
esrj-89750	237	33	belonging	belong	VERB
esrj-89750	237	34	category	category	NOUN
esrj-89750	237	35	of	of	ADP
esrj-89750	237	36	data	datum	NOUN
esrj-89750	237	37	,	,	PUNCT
esrj-89750	237	38	and	and	CCONJ
esrj-89750	237	39	the	the	DET
esrj-89750	237	40	total	total	ADJ
esrj-89750	237	41	number	number	NOUN
esrj-89750	237	42	of	of	ADP
esrj-89750	237	43	data	datum	NOUN
esrj-89750	237	44	in	in	ADP
esrj-89750	237	45	each	each	DET
esrj-89750	237	46	row	row	NOUN
esrj-89750	237	47	represents	represent	VERB
esrj-89750	237	48	the	the	DET
esrj-89750	237	49	number	number	NOUN
esrj-89750	237	50	of	of	ADP
esrj-89750	237	51	data	datum	NOUN
esrj-89750	237	52	instances	instance	NOUN
esrj-89750	237	53	in	in	ADP
esrj-89750	237	54	that	that	DET
esrj-89750	237	55	category	category	NOUN
esrj-89750	237	56	.	.	PUNCT
esrj-89750	238	1	the	the	DET
esrj-89750	238	2	sum	sum	NOUN
esrj-89750	238	3	of	of	ADP
esrj-89750	238	4	each	each	DET
esrj-89750	238	5	row	row	NOUN
esrj-89750	238	6	represents	represent	VERB
esrj-89750	238	7	the	the	DET
esrj-89750	238	8	true	true	ADJ
esrj-89750	238	9	number	number	NOUN
esrj-89750	238	10	of	of	ADP
esrj-89750	238	11	samples	sample	NOUN
esrj-89750	238	12	for	for	ADP
esrj-89750	238	13	the	the	DET
esrj-89750	238	14	category	category	NOUN
esrj-89750	238	15	,	,	PUNCT
esrj-89750	238	16	and	and	CCONJ
esrj-89750	238	17	the	the	DET
esrj-89750	238	18	sum	sum	NOUN
esrj-89750	238	19	of	of	ADP
esrj-89750	238	20	each	each	DET
esrj-89750	238	21	column	column	NOUN
esrj-89750	238	22	represents	represent	VERB
esrj-89750	238	23	the	the	DET
esrj-89750	238	24	number	number	NOUN
esrj-89750	238	25	of	of	ADP
esrj-89750	238	26	samples	sample	NOUN
esrj-89750	238	27	predicted	predict	VERB
esrj-89750	238	28	for	for	ADP
esrj-89750	238	29	the	the	DET
esrj-89750	238	30	category	category	NOUN
esrj-89750	238	31	.	.	PUNCT
esrj-89750	239	1	there	there	PRON
esrj-89750	239	2	are	be	VERB
esrj-89750	239	3	40	40	NUM
esrj-89750	239	4	test	test	NOUN
esrj-89750	239	5	sample	sample	NOUN
esrj-89750	239	6	data	datum	NOUN
esrj-89750	239	7	in	in	ADP
esrj-89750	239	8	this	this	DET
esrj-89750	239	9	experiment	experiment	NOUN
esrj-89750	239	10	,	,	PUNCT
esrj-89750	239	11	which	which	PRON
esrj-89750	239	12	are	be	AUX
esrj-89750	239	13	predicted	predict	VERB
esrj-89750	239	14	to	to	PART
esrj-89750	239	15	be	be	AUX
esrj-89750	239	16	four	four	NUM
esrj-89750	239	17	types	type	NOUN
esrj-89750	239	18	(	(	PUNCT
esrj-89750	239	19	chernozem	chernozem	NOUN
esrj-89750	239	20	,	,	PUNCT
esrj-89750	239	21	meadow	meadow	NOUN
esrj-89750	239	22	soil	soil	NOUN
esrj-89750	239	23	,	,	PUNCT
esrj-89750	239	24	marsh	marsh	ADJ
esrj-89750	239	25	soil	soil	NOUN
esrj-89750	239	26	and	and	CCONJ
esrj-89750	239	27	aeolian	aeolian	ADJ
esrj-89750	239	28	sandy	sandy	ADJ
esrj-89750	239	29	soil	soil	NOUN
esrj-89750	239	30	)	)	PUNCT
esrj-89750	239	31	,	,	PUNCT
esrj-89750	239	32	each	each	PRON
esrj-89750	239	33	of	of	ADP
esrj-89750	239	34	which	which	PRON
esrj-89750	239	35	has	have	VERB
esrj-89750	239	36	20	20	NUM
esrj-89750	239	37	samples	sample	NOUN
esrj-89750	239	38	.	.	PUNCT
esrj-89750	240	1	after	after	SCONJ
esrj-89750	240	2	the	the	DET
esrj-89750	240	3	three	three	NUM
esrj-89750	240	4	methods	method	NOUN
esrj-89750	240	5	are	be	AUX
esrj-89750	240	6	classified	classify	VERB
esrj-89750	240	7	,	,	PUNCT
esrj-89750	240	8	the	the	DET
esrj-89750	240	9	confusion	confusion	NOUN
esrj-89750	240	10	matrix	matrix	NOUN
esrj-89750	240	11	is	be	AUX
esrj-89750	240	12	shown	show	VERB
esrj-89750	240	13	in	in	ADP
esrj-89750	240	14	tables	table	NOUN
esrj-89750	240	15	4	4	NUM
esrj-89750	240	16	,	,	PUNCT
esrj-89750	240	17	5	5	NUM
esrj-89750	240	18	and	and	CCONJ
esrj-89750	240	19	6	6	NUM
esrj-89750	240	20	.	.	SYM
esrj-89750	240	21	364	364	NUM
esrj-89750	240	22	shujun	shujun	PROPN
esrj-89750	240	23	liang1	liang1	PROPN
esrj-89750	240	24	,	,	PUNCT
esrj-89750	240	25	jing	je	VERB
esrj-89750	241	1	cheng2	cheng2	PROPN
esrj-89750	241	2	*	*	PROPN
esrj-89750	241	3	,	,	PUNCT
esrj-89750	241	4	jianwei	jianwei	PROPN
esrj-89750	241	5	zhang1	zhang1	NOUN
esrj-89750	241	6	table	table	NOUN
esrj-89750	241	7	4	4	NUM
esrj-89750	241	8	.	.	PUNCT
esrj-89750	241	9	confusion	confusion	NOUN
esrj-89750	241	10	matrix	matrix	NOUN
esrj-89750	241	11	of	of	ADP
esrj-89750	241	12	the	the	DET
esrj-89750	241	13	classification	classification	NOUN
esrj-89750	241	14	results	result	NOUN
esrj-89750	241	15	of	of	ADP
esrj-89750	241	16	the	the	DET
esrj-89750	241	17	proposed	propose	VERB
esrj-89750	241	18	method	method	NOUN
esrj-89750	241	19	project	project	NOUN
esrj-89750	241	20	prediction	prediction	NOUN
esrj-89750	241	21	classification	classification	NOUN
esrj-89750	241	22	black	black	ADJ
esrj-89750	241	23	calcium	calcium	NOUN
esrj-89750	241	24	soil	soil	NOUN
esrj-89750	241	25	meadow	meadow	NOUN
esrj-89750	241	26	soil	soil	NOUN
esrj-89750	241	27	swamp	swamp	NOUN
esrj-89750	241	28	soil	soil	NOUN
esrj-89750	241	29	aeolian	aeolian	ADJ
esrj-89750	241	30	sand	sand	NOUN
esrj-89750	241	31	soil	soil	NOUN
esrj-89750	241	32	actual	actual	ADJ
esrj-89750	241	33	classification	classification	NOUN
esrj-89750	241	34	chernozem	chernozem	NOUN
esrj-89750	241	35	20	20	NUM
esrj-89750	241	36	0	0	NUM
esrj-89750	241	37	0	0	NUM
esrj-89750	241	38	0	0	NUM
esrj-89750	241	39	meadow	meadow	NOUN
esrj-89750	241	40	soil	soil	NOUN
esrj-89750	241	41	0	0	NUM
esrj-89750	241	42	20	20	NUM
esrj-89750	241	43	0	0	NUM
esrj-89750	241	44	1	1	NUM
esrj-89750	241	45	marsh	marsh	ADJ
esrj-89750	241	46	soil	soil	NOUN
esrj-89750	241	47	0	0	NUM
esrj-89750	241	48	0	0	NUM
esrj-89750	241	49	20	20	NUM
esrj-89750	241	50	0	0	NUM
esrj-89750	241	51	aeolian	aeolian	ADJ
esrj-89750	241	52	sandy	sandy	ADJ
esrj-89750	241	53	soil	soil	NOUN
esrj-89750	241	54	0	0	NUM
esrj-89750	241	55	0	0	NUM
esrj-89750	241	56	0	0	NUM
esrj-89750	241	57	19	19	NUM
esrj-89750	241	58	table	table	NOUN
esrj-89750	241	59	5	5	NUM
esrj-89750	241	60	.	.	PUNCT
esrj-89750	241	61	confusion	confusion	NOUN
esrj-89750	241	62	matrix	matrix	NOUN
esrj-89750	241	63	of	of	ADP
esrj-89750	241	64	classification	classification	NOUN
esrj-89750	241	65	results	result	NOUN
esrj-89750	241	66	of	of	ADP
esrj-89750	241	67	soil	soil	NOUN
esrj-89750	241	68	remote	remote	ADJ
esrj-89750	241	69	sensing	sense	VERB
esrj-89750	241	70	image	image	NOUN
esrj-89750	241	71	classification	classification	NOUN
esrj-89750	241	72	method	method	NOUN
esrj-89750	241	73	based	base	VERB
esrj-89750	241	74	on	on	ADP
esrj-89750	241	75	deep	deep	ADJ
esrj-89750	241	76	learning	learning	NOUN
esrj-89750	241	77	project	project	NOUN
esrj-89750	241	78	prediction	prediction	NOUN
esrj-89750	241	79	classification	classification	NOUN
esrj-89750	241	80	black	black	ADJ
esrj-89750	241	81	calcium	calcium	NOUN
esrj-89750	241	82	soil	soil	NOUN
esrj-89750	241	83	,	,	PUNCT
esrj-89750	241	84	meadow	meadow	NOUN
esrj-89750	241	85	soil	soil	NOUN
esrj-89750	241	86	swamp	swamp	NOUN
esrj-89750	241	87	soil	soil	NOUN
esrj-89750	241	88	aeolian	aeolian	ADJ
esrj-89750	241	89	sand	sand	NOUN
esrj-89750	241	90	soil	soil	NOUN
esrj-89750	241	91	actual	actual	ADJ
esrj-89750	241	92	classification	classification	NOUN
esrj-89750	241	93	chernozem	chernozem	NOUN
esrj-89750	241	94	17	17	NUM
esrj-89750	241	95	2	2	NUM
esrj-89750	241	96	1	1	NUM
esrj-89750	241	97	0	0	NUM
esrj-89750	241	98	meadow	meadow	NOUN
esrj-89750	241	99	soil	soil	NOUN
esrj-89750	241	100	0	0	NUM
esrj-89750	241	101	15	15	NUM
esrj-89750	241	102	1	1	NUM
esrj-89750	241	103	1	1	NUM
esrj-89750	241	104	marsh	marsh	NOUN
esrj-89750	241	105	soil	soil	NOUN
esrj-89750	241	106	2	2	NUM
esrj-89750	241	107	2	2	NUM
esrj-89750	241	108	16	16	NUM
esrj-89750	241	109	3	3	NUM
esrj-89750	241	110	aeolian	aeolian	ADJ
esrj-89750	241	111	sandy	sandy	ADJ
esrj-89750	241	112	soil	soil	NOUN
esrj-89750	241	113	1	1	NUM
esrj-89750	241	114	1	1	NUM
esrj-89750	241	115	2	2	NUM
esrj-89750	241	116	14	14	NUM
esrj-89750	241	117	table	table	NOUN
esrj-89750	241	118	6	6	NUM
esrj-89750	241	119	.	.	PUNCT
esrj-89750	241	120	confusion	confusion	NOUN
esrj-89750	241	121	matrix	matrix	NOUN
esrj-89750	241	122	of	of	ADP
esrj-89750	241	123	classification	classification	NOUN
esrj-89750	241	124	results	result	NOUN
esrj-89750	241	125	of	of	ADP
esrj-89750	241	126	soil	soil	NOUN
esrj-89750	241	127	remote	remote	ADJ
esrj-89750	241	128	sensing	sense	VERB
esrj-89750	241	129	image	image	NOUN
esrj-89750	241	130	classification	classification	NOUN
esrj-89750	241	131	method	method	NOUN
esrj-89750	241	132	based	base	VERB
esrj-89750	241	133	on	on	ADP
esrj-89750	241	134	maximum	maximum	ADJ
esrj-89750	241	135	likelihood	likelihood	NOUN
esrj-89750	241	136	estimation	estimation	NOUN
esrj-89750	241	137	project	project	NOUN
esrj-89750	241	138	prediction	prediction	NOUN
esrj-89750	241	139	classification	classification	NOUN
esrj-89750	241	140	black	black	ADJ
esrj-89750	241	141	calcium	calcium	NOUN
esrj-89750	241	142	soil	soil	NOUN
esrj-89750	241	143	meadow	meadow	NOUN
esrj-89750	241	144	soil	soil	NOUN
esrj-89750	241	145	swamp	swamp	NOUN
esrj-89750	241	146	soil	soil	NOUN
esrj-89750	241	147	aeolian	aeolian	ADJ
esrj-89750	241	148	sand	sand	NOUN
esrj-89750	241	149	soil	soil	NOUN
esrj-89750	241	150	actual	actual	ADJ
esrj-89750	241	151	classification	classification	NOUN
esrj-89750	241	152	chernozem	chernozem	NOUN
esrj-89750	241	153	18	18	NUM
esrj-89750	241	154	2	2	NUM
esrj-89750	241	155	1	1	NUM
esrj-89750	241	156	1	1	NUM
esrj-89750	241	157	meadow	meadow	NOUN
esrj-89750	241	158	soil	soil	NOUN
esrj-89750	241	159	0	0	NUM
esrj-89750	241	160	14	14	NUM
esrj-89750	241	161	1	1	NUM
esrj-89750	241	162	1	1	NUM
esrj-89750	242	1	marsh	marsh	NOUN
esrj-89750	242	2	soil	soil	NOUN
esrj-89750	242	3	1	1	NUM
esrj-89750	242	4	3	3	NUM
esrj-89750	242	5	18	18	NUM
esrj-89750	242	6	3	3	NUM
esrj-89750	242	7	aeolian	aeolian	ADJ
esrj-89750	242	8	sandy	sandy	ADJ
esrj-89750	242	9	soil	soil	NOUN
esrj-89750	242	10	1	1	NUM
esrj-89750	242	11	1	1	NUM
esrj-89750	242	12	0	0	NUM
esrj-89750	242	13	15	15	NUM
esrj-89750	242	14	comparing	compare	VERB
esrj-89750	242	15	tables	table	NOUN
esrj-89750	242	16	4	4	NUM
esrj-89750	242	17	,	,	PUNCT
esrj-89750	242	18	5	5	NUM
esrj-89750	242	19	and	and	CCONJ
esrj-89750	242	20	6	6	NUM
esrj-89750	242	21	,	,	PUNCT
esrj-89750	242	22	it	it	PRON
esrj-89750	242	23	can	can	AUX
esrj-89750	242	24	be	be	AUX
esrj-89750	242	25	seen	see	VERB
esrj-89750	242	26	that	that	SCONJ
esrj-89750	242	27	after	after	SCONJ
esrj-89750	242	28	the	the	DET
esrj-89750	242	29	proposed	propose	VERB
esrj-89750	242	30	method	method	NOUN
esrj-89750	242	31	is	be	AUX
esrj-89750	242	32	used	use	VERB
esrj-89750	242	33	to	to	PART
esrj-89750	242	34	classify	classify	VERB
esrj-89750	242	35	,	,	PUNCT
esrj-89750	242	36	20	20	NUM
esrj-89750	242	37	samples	sample	NOUN
esrj-89750	242	38	in	in	ADP
esrj-89750	242	39	the	the	DET
esrj-89750	242	40	first	first	ADJ
esrj-89750	242	41	,	,	PUNCT
esrj-89750	242	42	second	second	ADJ
esrj-89750	242	43	and	and	CCONJ
esrj-89750	242	44	third	third	ADJ
esrj-89750	242	45	rows	row	NOUN
esrj-89750	242	46	of	of	ADP
esrj-89750	242	47	the	the	DET
esrj-89750	242	48	confusion	confusion	NOUN
esrj-89750	242	49	matrix	matrix	NOUN
esrj-89750	242	50	established	establish	VERB
esrj-89750	242	51	according	accord	VERB
esrj-89750	242	52	to	to	ADP
esrj-89750	242	53	the	the	DET
esrj-89750	242	54	results	result	NOUN
esrj-89750	242	55	correspond	correspond	VERB
esrj-89750	242	56	to	to	ADP
esrj-89750	242	57	the	the	DET
esrj-89750	242	58	first	first	ADJ
esrj-89750	242	59	,	,	PUNCT
esrj-89750	242	60	second	second	ADJ
esrj-89750	242	61	and	and	CCONJ
esrj-89750	242	62	third	third	ADJ
esrj-89750	242	63	categories	category	NOUN
esrj-89750	242	64	respectively	respectively	ADV
esrj-89750	242	65	,	,	PUNCT
esrj-89750	242	66	indicating	indicate	VERB
esrj-89750	242	67	that	that	SCONJ
esrj-89750	242	68	the	the	DET
esrj-89750	242	69	samples	sample	NOUN
esrj-89750	242	70	are	be	AUX
esrj-89750	242	71	all	all	PRON
esrj-89750	242	72	correctly	correctly	ADV
esrj-89750	242	73	predicted	predict	VERB
esrj-89750	242	74	,	,	PUNCT
esrj-89750	242	75	and	and	CCONJ
esrj-89750	242	76	only	only	ADV
esrj-89750	242	77	one	one	NUM
esrj-89750	242	78	sample	sample	NOUN
esrj-89750	242	79	belonging	belong	VERB
esrj-89750	242	80	to	to	ADP
esrj-89750	242	81	the	the	DET
esrj-89750	242	82	fourth	fourth	ADJ
esrj-89750	242	83	category	category	NOUN
esrj-89750	242	84	in	in	ADP
esrj-89750	242	85	the	the	DET
esrj-89750	242	86	fourth	fourth	ADJ
esrj-89750	242	87	row	row	NOUN
esrj-89750	242	88	is	be	AUX
esrj-89750	242	89	misclassified	misclassifie	VERB
esrj-89750	242	90	into	into	ADP
esrj-89750	242	91	the	the	DET
esrj-89750	242	92	second	second	ADJ
esrj-89750	242	93	category	category	NOUN
esrj-89750	242	94	.	.	PUNCT
esrj-89750	243	1	this	this	DET
esrj-89750	243	2	result	result	NOUN
esrj-89750	243	3	is	be	AUX
esrj-89750	243	4	much	much	ADV
esrj-89750	243	5	better	well	ADJ
esrj-89750	243	6	than	than	ADP
esrj-89750	243	7	the	the	DET
esrj-89750	243	8	other	other	ADJ
esrj-89750	243	9	two	two	NUM
esrj-89750	243	10	methods	method	NOUN
esrj-89750	243	11	,	,	PUNCT
esrj-89750	243	12	which	which	PRON
esrj-89750	243	13	shows	show	VERB
esrj-89750	243	14	that	that	SCONJ
esrj-89750	243	15	the	the	DET
esrj-89750	243	16	performance	performance	NOUN
esrj-89750	243	17	of	of	ADP
esrj-89750	243	18	the	the	DET
esrj-89750	243	19	proposed	propose	VERB
esrj-89750	243	20	method	method	NOUN
esrj-89750	243	21	is	be	AUX
esrj-89750	243	22	better	well	ADJ
esrj-89750	243	23	.	.	PUNCT
esrj-89750	244	1	conclusions	conclusion	NOUN
esrj-89750	244	2	in	in	ADP
esrj-89750	244	3	summary	summary	NOUN
esrj-89750	244	4	,	,	PUNCT
esrj-89750	244	5	in	in	ADP
esrj-89750	244	6	order	order	NOUN
esrj-89750	244	7	to	to	PART
esrj-89750	244	8	give	give	VERB
esrj-89750	244	9	full	full	ADJ
esrj-89750	244	10	play	play	NOUN
esrj-89750	244	11	to	to	ADP
esrj-89750	244	12	the	the	DET
esrj-89750	244	13	advantages	advantage	NOUN
esrj-89750	244	14	of	of	ADP
esrj-89750	244	15	the	the	DET
esrj-89750	244	16	maximum	maximum	ADJ
esrj-89750	244	17	likelihood	likelihood	NOUN
esrj-89750	244	18	method	method	NOUN
esrj-89750	244	19	,	,	PUNCT
esrj-89750	244	20	a	a	DET
esrj-89750	244	21	classification	classification	NOUN
esrj-89750	244	22	method	method	NOUN
esrj-89750	244	23	for	for	ADP
esrj-89750	244	24	making	make	VERB
esrj-89750	244	25	best	good	ADJ
esrj-89750	244	26	use	use	NOUN
esrj-89750	244	27	of	of	ADP
esrj-89750	244	28	the	the	DET
esrj-89750	244	29	advantages	advantage	NOUN
esrj-89750	244	30	and	and	CCONJ
esrj-89750	244	31	bypass	bypass	VERB
esrj-89750	244	32	the	the	DET
esrj-89750	244	33	disadvantages	disadvantage	NOUN
esrj-89750	244	34	is	be	AUX
esrj-89750	244	35	designed	design	VERB
esrj-89750	244	36	,	,	PUNCT
esrj-89750	244	37	that	that	ADV
esrj-89750	244	38	is	is	ADV
esrj-89750	244	39	,	,	PUNCT
esrj-89750	244	40	the	the	DET
esrj-89750	244	41	maximum	maximum	ADJ
esrj-89750	244	42	likelihood	likelihood	NOUN
esrj-89750	244	43	classification	classification	NOUN
esrj-89750	244	44	method	method	NOUN
esrj-89750	244	45	for	for	ADP
esrj-89750	244	46	soil	soil	NOUN
esrj-89750	244	47	remote	remote	ADJ
esrj-89750	244	48	sensing	sensing	NOUN
esrj-89750	244	49	images	image	NOUN
esrj-89750	244	50	based	base	VERB
esrj-89750	244	51	on	on	ADP
esrj-89750	244	52	deep	deep	ADJ
esrj-89750	244	53	learning	learning	NOUN
esrj-89750	244	54	.	.	PUNCT
esrj-89750	245	1	firstly	firstly	ADV
esrj-89750	245	2	,	,	PUNCT
esrj-89750	245	3	deep	deep	ADJ
esrj-89750	245	4	sensing	sensing	NOUN
esrj-89750	245	5	is	be	AUX
esrj-89750	245	6	used	use	VERB
esrj-89750	245	7	to	to	PART
esrj-89750	245	8	detect	detect	VERB
esrj-89750	245	9	remote	remote	ADJ
esrj-89750	245	10	sensing	sense	VERB
esrj-89750	245	11	image	image	NOUN
esrj-89750	245	12	targets	target	NOUN
esrj-89750	245	13	and	and	CCONJ
esrj-89750	245	14	extract	extract	VERB
esrj-89750	245	15	the	the	DET
esrj-89750	245	16	target	target	NOUN
esrj-89750	245	17	classification	classification	NOUN
esrj-89750	245	18	features	feature	NOUN
esrj-89750	245	19	of	of	ADP
esrj-89750	245	20	multiple	multiple	ADJ
esrj-89750	245	21	kinds	kind	NOUN
esrj-89750	245	22	of	of	ADP
esrj-89750	245	23	images	image	NOUN
esrj-89750	245	24	;	;	PUNCT
esrj-89750	245	25	then	then	ADV
esrj-89750	245	26	using	use	VERB
esrj-89750	245	27	the	the	DET
esrj-89750	245	28	maximum	maximum	ADJ
esrj-89750	245	29	likelihood	likelihood	NOUN
esrj-89750	245	30	estimation	estimation	NOUN
esrj-89750	245	31	algorithm	algorithm	NOUN
esrj-89750	245	32	,	,	PUNCT
esrj-89750	245	33	soil	soil	NOUN
esrj-89750	245	34	remote	remote	ADJ
esrj-89750	245	35	sensing	sensing	NOUN
esrj-89750	245	36	images	image	NOUN
esrj-89750	245	37	are	be	AUX
esrj-89750	245	38	to	to	PART
esrj-89750	245	39	classify	classify	VERB
esrj-89750	245	40	.	.	PUNCT
esrj-89750	246	1	finally	finally	ADV
esrj-89750	246	2	,	,	PUNCT
esrj-89750	246	3	the	the	DET
esrj-89750	246	4	experimental	experimental	ADJ
esrj-89750	246	5	results	result	NOUN
esrj-89750	246	6	show	show	VERB
esrj-89750	246	7	that	that	SCONJ
esrj-89750	246	8	the	the	DET
esrj-89750	246	9	proposed	propose	VERB
esrj-89750	246	10	method	method	NOUN
esrj-89750	246	11	not	not	PART
esrj-89750	246	12	only	only	ADV
esrj-89750	246	13	can	can	AUX
esrj-89750	246	14	realize	realize	VERB
esrj-89750	246	15	the	the	DET
esrj-89750	246	16	classification	classification	NOUN
esrj-89750	246	17	of	of	ADP
esrj-89750	246	18	remote	remote	ADJ
esrj-89750	246	19	sensing	sense	VERB
esrj-89750	246	20	data	datum	NOUN
esrj-89750	246	21	,	,	PUNCT
esrj-89750	246	22	but	but	CCONJ
esrj-89750	246	23	also	also	ADV
esrj-89750	246	24	has	have	VERB
esrj-89750	246	25	a	a	DET
esrj-89750	246	26	higher	high	ADJ
esrj-89750	246	27	overall	overall	ADJ
esrj-89750	246	28	classification	classification	NOUN
esrj-89750	246	29	accuracy	accuracy	NOUN
esrj-89750	246	30	.	.	PUNCT
esrj-89750	247	1	compared	compare	VERB
esrj-89750	247	2	with	with	ADP
esrj-89750	247	3	the	the	DET
esrj-89750	247	4	two	two	NUM
esrj-89750	247	5	single	single	ADJ
esrj-89750	247	6	classification	classification	NOUN
esrj-89750	247	7	methods	method	NOUN
esrj-89750	247	8	,	,	PUNCT
esrj-89750	247	9	the	the	DET
esrj-89750	247	10	classification	classification	NOUN
esrj-89750	247	11	results	result	NOUN
esrj-89750	247	12	obtained	obtain	VERB
esrj-89750	247	13	by	by	ADP
esrj-89750	247	14	the	the	DET
esrj-89750	247	15	proposed	propose	VERB
esrj-89750	247	16	method	method	NOUN
esrj-89750	247	17	are	be	AUX
esrj-89750	247	18	more	more	ADV
esrj-89750	247	19	accurate	accurate	ADJ
esrj-89750	247	20	.	.	PUNCT
esrj-89750	248	1	acknowledgement	acknowledgement	NOUN
esrj-89750	248	2	this	this	DET
esrj-89750	248	3	work	work	NOUN
esrj-89750	248	4	was	be	AUX
esrj-89750	248	5	supported	support	VERB
esrj-89750	248	6	by	by	ADP
esrj-89750	248	7	research	research	NOUN
esrj-89750	248	8	on	on	ADP
esrj-89750	248	9	key	key	ADJ
esrj-89750	248	10	technology	technology	NOUN
esrj-89750	248	11	of	of	ADP
esrj-89750	248	12	disease	disease	NOUN
esrj-89750	248	13	biomarker	biomarker	NOUN
esrj-89750	248	14	identification	identification	NOUN
esrj-89750	248	15	based	base	VERB
esrj-89750	248	16	on	on	ADP
esrj-89750	248	17	deep	deep	ADJ
esrj-89750	248	18	learning	learning	NOUN
esrj-89750	248	19	,	,	PUNCT
esrj-89750	248	20	a	a	DET
esrj-89750	248	21	grant	grant	NOUN
esrj-89750	248	22	from	from	ADP
esrj-89750	248	23	key	key	ADJ
esrj-89750	248	24	r&d	r&d	NOUN
esrj-89750	248	25	and	and	CCONJ
esrj-89750	248	26	promotion	promotion	NOUN
esrj-89750	248	27	projects	project	NOUN
esrj-89750	248	28	in	in	ADP
esrj-89750	248	29	henan	henan	PROPN
esrj-89750	248	30	province	province	PROPN
esrj-89750	248	31	of	of	ADP
esrj-89750	248	32	china	china	PROPN
esrj-89750	248	33	(	(	PUNCT
esrj-89750	248	34	grant	grant	PROPN
esrj-89750	248	35	no.192102210132	no.192102210132	PROPN
esrj-89750	248	36	)	)	PUNCT
esrj-89750	248	37	.	.	PUNCT
esrj-89750	249	1	references	reference	NOUN
esrj-89750	249	2	cheng	cheng	PROPN
esrj-89750	249	3	l.	l.	PROPN
esrj-89750	249	4	y.	y.	PROPN
esrj-89750	249	5	,	,	PUNCT
esrj-89750	249	6	meng	meng	PROPN
esrj-89750	249	7	x.	x.	PROPN
esrj-89750	249	8	y.	y.	PROPN
esrj-89750	249	9	,	,	PUNCT
esrj-89750	249	10	&	&	CCONJ
esrj-89750	249	11	da	da	PROPN
esrj-89750	249	12	,	,	PUNCT
esrj-89750	249	13	x.	x.	NOUN
esrj-89750	249	14	m.	m.	NOUN
esrj-89750	249	15	(	(	PUNCT
esrj-89750	249	16	2018	2018	NUM
esrj-89750	249	17	)	)	PUNCT
esrj-89750	249	18	.	.	PUNCT
esrj-89750	250	1	simulation	simulation	NOUN
esrj-89750	250	2	of	of	ADP
esrj-89750	250	3	efficient	efficient	ADJ
esrj-89750	250	4	classification	classification	NOUN
esrj-89750	250	5	for	for	ADP
esrj-89750	250	6	image	image	NOUN
esrj-89750	250	7	characteristics	characteristic	NOUN
esrj-89750	250	8	of	of	ADP
esrj-89750	250	9	fruit	fruit	NOUN
esrj-89750	250	10	in	in	ADP
esrj-89750	250	11	fruit	fruit	NOUN
esrj-89750	250	12	pest	pest	NOUN
esrj-89750	250	13	.	.	PUNCT
esrj-89750	251	1	computer	computer	NOUN
esrj-89750	251	2	simulation	simulation	NOUN
esrj-89750	251	3	,	,	PUNCT
esrj-89750	251	4	35(02):425	35(02):425	NUM
esrj-89750	251	5	-	-	SYM
esrj-89750	251	6	428	428	NUM
esrj-89750	251	7	.	.	PUNCT
esrj-89750	252	1	chen	chen	PROPN
esrj-89750	252	2	,	,	PUNCT
esrj-89750	252	3	w.	w.	PROPN
esrj-89750	252	4	,	,	PUNCT
esrj-89750	252	5	xie	xie	PROPN
esrj-89750	252	6	,	,	PUNCT
esrj-89750	252	7	z.	z.	PROPN
esrj-89750	252	8	,	,	PUNCT
esrj-89750	252	9	ma	ma	PROPN
esrj-89750	252	10	,	,	PUNCT
esrj-89750	252	11	l.	l.	PROPN
esrj-89750	252	12	,	,	PUNCT
esrj-89750	252	13	liu	liu	PROPN
esrj-89750	252	14	,	,	PUNCT
esrj-89750	252	15	j.	j.	PROPN
esrj-89750	252	16	,	,	PUNCT
esrj-89750	252	17	&	&	CCONJ
esrj-89750	252	18	liang	liang	PROPN
esrj-89750	252	19	,	,	PUNCT
esrj-89750	252	20	x.	x.	NOUN
esrj-89750	252	21	(	(	PUNCT
esrj-89750	252	22	2019	2019	NUM
esrj-89750	252	23	)	)	PUNCT
esrj-89750	252	24	.	.	PUNCT
esrj-89750	253	1	a	a	DET
esrj-89750	253	2	faster	fast	ADJ
esrj-89750	253	3	maximumlikelihood	maximumlikelihood	NOUN
esrj-89750	253	4	modulation	modulation	NOUN
esrj-89750	253	5	classification	classification	NOUN
esrj-89750	253	6	in	in	ADP
esrj-89750	253	7	flat	flat	ADJ
esrj-89750	253	8	fading	fade	VERB
esrj-89750	253	9	non	non	ADJ
esrj-89750	253	10	-	-	ADJ
esrj-89750	253	11	gaussian	gaussian	ADJ
esrj-89750	253	12	channels	channel	NOUN
esrj-89750	253	13	.	.	PUNCT
esrj-89750	254	1	ieee	ieee	NOUN
esrj-89750	254	2	communications	communication	NOUN
esrj-89750	254	3	letters	letter	NOUN
esrj-89750	254	4	,	,	PUNCT
esrj-89750	254	5	23(3):454	23(3):454	PROPN
esrj-89750	254	6	-	-	SYM
esrj-89750	254	7	457	457	NUM
esrj-89750	254	8	.	.	PUNCT
esrj-89750	255	1	demattê	demattê	PROPN
esrj-89750	255	2	,	,	PUNCT
esrj-89750	255	3	j.	j.	PROPN
esrj-89750	255	4	a.	a.	PROPN
esrj-89750	255	5	m.	m.	PROPN
esrj-89750	255	6	,	,	PUNCT
esrj-89750	255	7	sayão	sayão	NOUN
esrj-89750	255	8	,	,	PUNCT
esrj-89750	255	9	v.	v.	ADP
esrj-89750	255	10	m.	m.	NOUN
esrj-89750	255	11	,	,	PUNCT
esrj-89750	255	12	rizzo	rizzo	PROPN
esrj-89750	255	13	,	,	PUNCT
esrj-89750	255	14	r.	r.	PROPN
esrj-89750	255	15	,	,	PUNCT
esrj-89750	255	16	&	&	CCONJ
esrj-89750	255	17	fongaro	fongaro	PROPN
esrj-89750	255	18	,	,	PUNCT
esrj-89750	255	19	c.	c.	PROPN
esrj-89750	255	20	t.	t.	PROPN
esrj-89750	255	21	(	(	PUNCT
esrj-89750	255	22	2017	2017	NUM
esrj-89750	255	23	)	)	PUNCT
esrj-89750	255	24	.	.	PUNCT
esrj-89750	256	1	soil	soil	NOUN
esrj-89750	256	2	class	class	NOUN
esrj-89750	256	3	and	and	CCONJ
esrj-89750	256	4	attribute	attribute	NOUN
esrj-89750	256	5	dynamics	dynamic	NOUN
esrj-89750	256	6	and	and	CCONJ
esrj-89750	256	7	their	their	PRON
esrj-89750	256	8	relationship	relationship	NOUN
esrj-89750	256	9	with	with	ADP
esrj-89750	256	10	natural	natural	ADJ
esrj-89750	256	11	vegetation	vegetation	NOUN
esrj-89750	256	12	based	base	VERB
esrj-89750	256	13	on	on	ADP
esrj-89750	256	14	satellite	satellite	NOUN
esrj-89750	256	15	remote	remote	ADJ
esrj-89750	256	16	sensing	sensing	NOUN
esrj-89750	256	17	.	.	PUNCT
esrj-89750	257	1	geoderma	geoderma	PROPN
esrj-89750	257	2	,	,	PUNCT
esrj-89750	257	3	302:39	302:39	PROPN
esrj-89750	257	4	-	-	SYM
esrj-89750	257	5	51	51	NUM
esrj-89750	257	6	.	.	PUNCT
esrj-89750	258	1	fitak	fitak	PROPN
esrj-89750	258	2	,	,	PUNCT
esrj-89750	258	3	r.	r.	PROPN
esrj-89750	258	4	r.	r.	PROPN
esrj-89750	258	5	,	,	PUNCT
esrj-89750	258	6	&	&	CCONJ
esrj-89750	258	7	johnsen	johnsen	PROPN
esrj-89750	258	8	,	,	PUNCT
esrj-89750	258	9	s.	s.	PROPN
esrj-89750	258	10	(	(	PUNCT
esrj-89750	258	11	2017	2017	NUM
esrj-89750	258	12	)	)	PUNCT
esrj-89750	258	13	.	.	PUNCT
esrj-89750	259	1	bringing	bring	VERB
esrj-89750	259	2	the	the	DET
esrj-89750	259	3	analysis	analysis	NOUN
esrj-89750	259	4	of	of	ADP
esrj-89750	259	5	animal	animal	NOUN
esrj-89750	259	6	orientation	orientation	NOUN
esrj-89750	259	7	data	datum	NOUN
esrj-89750	259	8	full	full	ADJ
esrj-89750	259	9	circle	circle	NOUN
esrj-89750	259	10	:	:	PUNCT
esrj-89750	259	11	model	model	NOUN
esrj-89750	259	12	-	-	PUNCT
esrj-89750	259	13	based	base	VERB
esrj-89750	259	14	approaches	approach	NOUN
esrj-89750	259	15	with	with	ADP
esrj-89750	259	16	maximum	maximum	ADJ
esrj-89750	259	17	likelihood	likelihood	NOUN
esrj-89750	259	18	.	.	PUNCT
esrj-89750	260	1	journal	journal	NOUN
esrj-89750	260	2	of	of	ADP
esrj-89750	260	3	experimental	experimental	ADJ
esrj-89750	260	4	biology	biology	NOUN
esrj-89750	260	5	,	,	PUNCT
esrj-89750	260	6	220(21):jeb.167056	220(21):jeb.167056	PROPN
esrj-89750	260	7	.	.	PUNCT
esrj-89750	260	8	handelman	handelman	NOUN
esrj-89750	260	9	,	,	PUNCT
esrj-89750	260	10	t.	t.	PROPN
esrj-89750	260	11	,	,	PUNCT
esrj-89750	260	12	&	&	CCONJ
esrj-89750	260	13	chor	chor	PROPN
esrj-89750	260	14	,	,	PUNCT
esrj-89750	260	15	b.	b.	PROPN
esrj-89750	260	16	(	(	PUNCT
esrj-89750	260	17	2017	2017	NUM
esrj-89750	260	18	)	)	PUNCT
esrj-89750	260	19	.	.	PUNCT
esrj-89750	261	1	cases	case	NOUN
esrj-89750	261	2	in	in	ADP
esrj-89750	261	3	which	which	PRON
esrj-89750	261	4	ancestral	ancestral	ADJ
esrj-89750	261	5	maximum	maximum	ADJ
esrj-89750	261	6	likelihood	likelihood	NOUN
esrj-89750	261	7	will	will	AUX
esrj-89750	261	8	be	be	AUX
esrj-89750	261	9	confusingly	confusingly	ADV
esrj-89750	261	10	misleading	misleading	ADJ
esrj-89750	261	11	.	.	PUNCT
esrj-89750	262	1	journal	journal	NOUN
esrj-89750	262	2	of	of	ADP
esrj-89750	262	3	theoretical	theoretical	ADJ
esrj-89750	262	4	biology	biology	NOUN
esrj-89750	262	5	,	,	PUNCT
esrj-89750	262	6	420:318	420:318	NOUN
esrj-89750	262	7	-	-	PUNCT
esrj-89750	262	8	323	323	NUM
esrj-89750	262	9	.	.	PUNCT
esrj-89750	263	1	he	he	PRON
esrj-89750	263	2	,	,	PUNCT
esrj-89750	263	3	j.	j.	PROPN
esrj-89750	263	4	,	,	PUNCT
esrj-89750	263	5	liu	liu	PROPN
esrj-89750	263	6	,	,	PUNCT
esrj-89750	263	7	g.	g.	PROPN
esrj-89750	263	8	,	,	PUNCT
esrj-89750	263	9	li	li	PROPN
esrj-89750	263	10	,	,	PUNCT
esrj-89750	263	11	w.	w.	PROPN
esrj-89750	263	12	,	,	PUNCT
esrj-89750	263	13	tang	tang	PROPN
esrj-89750	263	14	,	,	PUNCT
esrj-89750	263	15	c.	c.	PROPN
esrj-89750	263	16	,	,	PUNCT
esrj-89750	263	17	&	&	CCONJ
esrj-89750	263	18	lu	lu	PROPN
esrj-89750	263	19	,	,	PUNCT
esrj-89750	263	20	j.	j.	PROPN
esrj-89750	263	21	(	(	PUNCT
esrj-89750	263	22	2018	2018	NUM
esrj-89750	263	23	)	)	PUNCT
esrj-89750	263	24	.	.	PUNCT
esrj-89750	264	1	an	an	DET
esrj-89750	264	2	evaluation	evaluation	NOUN
esrj-89750	264	3	approach	approach	NOUN
esrj-89750	264	4	for	for	ADP
esrj-89750	264	5	segmentation	segmentation	NOUN
esrj-89750	264	6	results	result	NOUN
esrj-89750	264	7	of	of	ADP
esrj-89750	264	8	high	high	ADJ
esrj-89750	264	9	-	-	PUNCT
esrj-89750	264	10	resolution	resolution	NOUN
esrj-89750	264	11	remote	remote	ADJ
esrj-89750	264	12	sensing	sensing	NOUN
esrj-89750	264	13	images	image	NOUN
esrj-89750	264	14	based	base	VERB
esrj-89750	264	15	on	on	ADP
esrj-89750	264	16	the	the	DET
esrj-89750	264	17	degree	degree	NOUN
esrj-89750	264	18	distribution	distribution	NOUN
esrj-89750	264	19	of	of	ADP
esrj-89750	264	20	land	land	NOUN
esrj-89750	264	21	cover	cover	NOUN
esrj-89750	264	22	networks	network	NOUN
esrj-89750	264	23	.	.	PUNCT
esrj-89750	265	1	international	international	ADJ
esrj-89750	265	2	journal	journal	NOUN
esrj-89750	265	3	of	of	ADP
esrj-89750	265	4	modern	modern	ADJ
esrj-89750	265	5	physics	physics	PROPN
esrj-89750	265	6	b	b	PROPN
esrj-89750	265	7	,	,	PUNCT
esrj-89750	265	8	32(25):1850283	32(25):1850283	NUM
esrj-89750	265	9	.	.	PUNCT
esrj-89750	266	1	li	li	PROPN
esrj-89750	266	2	w	w	PROPN
esrj-89750	266	3	,	,	PUNCT
esrj-89750	266	4	hu	hu	PROPN
esrj-89750	266	5	x	x	PROPN
esrj-89750	266	6	,	,	PUNCT
esrj-89750	266	7	du	du	PROPN
esrj-89750	266	8	j	j	PROPN
esrj-89750	266	9	,	,	PUNCT
esrj-89750	266	10	et	et	PROPN
esrj-89750	266	11	al	al	PROPN
esrj-89750	266	12	(	(	PUNCT
esrj-89750	266	13	2017	2017	NUM
esrj-89750	266	14	)	)	PUNCT
esrj-89750	266	15	adaptive	adaptive	ADJ
esrj-89750	266	16	remote	remote	ADJ
esrj-89750	266	17	-	-	PUNCT
esrj-89750	266	18	sensing	sense	VERB
esrj-89750	266	19	image	image	NOUN
esrj-89750	266	20	fusion	fusion	NOUN
esrj-89750	266	21	based	base	VERB
esrj-89750	266	22	on	on	ADP
esrj-89750	266	23	dynamic	dynamic	ADJ
esrj-89750	266	24	gradient	gradient	NOUN
esrj-89750	266	25	sparse	sparse	ADJ
esrj-89750	266	26	and	and	CCONJ
esrj-89750	266	27	average	average	ADJ
esrj-89750	266	28	gradient	gradient	ADJ
esrj-89750	266	29	difference	difference	NOUN
esrj-89750	266	30	.	.	PUNCT
esrj-89750	267	1	international	international	ADJ
esrj-89750	267	2	journal	journal	NOUN
esrj-89750	267	3	of	of	ADP
esrj-89750	267	4	remote	remote	ADJ
esrj-89750	267	5	sensing	sense	VERB
esrj-89750	267	6	38(23):7316	38(23):7316	PROPN
esrj-89750	267	7	-	-	SYM
esrj-89750	267	8	7332	7332	NUM
esrj-89750	267	9	.	.	PUNCT
esrj-89750	268	1	ma	ma	PROPN
esrj-89750	268	2	,	,	PUNCT
esrj-89750	268	3	x.	x.	PROPN
esrj-89750	268	4	,	,	PUNCT
esrj-89750	268	5	liu	liu	PROPN
esrj-89750	268	6	,	,	PUNCT
esrj-89750	268	7	w.	w.	PROPN
esrj-89750	268	8	,	,	PUNCT
esrj-89750	268	9	li	li	PROPN
esrj-89750	268	10	,	,	PUNCT
esrj-89750	268	11	s.	s.	PROPN
esrj-89750	268	12	,	,	PUNCT
esrj-89750	268	13	&	&	CCONJ
esrj-89750	268	14	zhou	zhou	PROPN
esrj-89750	268	15	,	,	PUNCT
esrj-89750	268	16	y.	y.	PROPN
esrj-89750	268	17	(	(	PUNCT
esrj-89750	268	18	2018	2018	NUM
esrj-89750	268	19	)	)	PUNCT
esrj-89750	268	20	.	.	PUNCT
esrj-89750	269	1	hypergraph	hypergraph	VERB
esrj-89750	269	2	p	p	NOUN
esrj-89750	269	3	-	-	PUNCT
esrj-89750	269	4	laplacian	laplacian	ADJ
esrj-89750	269	5	regularization	regularization	NOUN
esrj-89750	269	6	for	for	ADP
esrj-89750	269	7	remote	remote	ADJ
esrj-89750	269	8	sensing	sense	VERB
esrj-89750	269	9	image	image	NOUN
esrj-89750	269	10	recognition	recognition	NOUN
esrj-89750	269	11	.	.	PUNCT
esrj-89750	270	1	ieee	ieee	NOUN
esrj-89750	270	2	transactions	transaction	NOUN
esrj-89750	270	3	on	on	ADP
esrj-89750	270	4	geoscience	geoscience	NOUN
esrj-89750	270	5	and	and	CCONJ
esrj-89750	270	6	remote	remote	ADJ
esrj-89750	270	7	sensing	sensing	NOUN
esrj-89750	270	8	,	,	PUNCT
esrj-89750	270	9	1	1	NUM
esrj-89750	270	10	-	-	SYM
esrj-89750	270	11	11	11	NUM
esrj-89750	270	12	.	.	PUNCT
esrj-89750	271	1	mccord	mccord	PROPN
esrj-89750	271	2	,	,	PUNCT
esrj-89750	271	3	s.	s.	PROPN
esrj-89750	271	4	e.	e.	PROPN
esrj-89750	271	5	,	,	PUNCT
esrj-89750	271	6	buenemann	buenemann	PROPN
esrj-89750	271	7	,	,	PUNCT
esrj-89750	271	8	m.	m.	NOUN
esrj-89750	271	9	,	,	PUNCT
esrj-89750	271	10	karl	karl	PROPN
esrj-89750	271	11	,	,	PUNCT
esrj-89750	271	12	j.	j.	PROPN
esrj-89750	271	13	w.	w.	PROPN
esrj-89750	271	14	,	,	PUNCT
esrj-89750	271	15	browning	browning	PROPN
esrj-89750	271	16	,	,	PUNCT
esrj-89750	271	17	d.	d.	PROPN
esrj-89750	271	18	m.	m.	PROPN
esrj-89750	271	19	,	,	PUNCT
esrj-89750	271	20	&	&	CCONJ
esrj-89750	271	21	hadley	hadley	PROPN
esrj-89750	271	22	,	,	PUNCT
esrj-89750	271	23	b.	b.	PROPN
esrj-89750	271	24	c.	c.	PROPN
esrj-89750	271	25	(	(	PUNCT
esrj-89750	271	26	2017	2017	NUM
esrj-89750	271	27	)	)	PUNCT
esrj-89750	271	28	.	.	PUNCT
esrj-89750	272	1	integrating	integrate	VERB
esrj-89750	272	2	remotely	remotely	ADV
esrj-89750	272	3	sensed	sense	VERB
esrj-89750	272	4	imagery	imagery	NOUN
esrj-89750	272	5	and	and	CCONJ
esrj-89750	272	6	existing	exist	VERB
esrj-89750	272	7	multiscale	multiscale	ADJ
esrj-89750	272	8	field	field	NOUN
esrj-89750	272	9	data	datum	NOUN
esrj-89750	272	10	to	to	PART
esrj-89750	272	11	derive	derive	VERB
esrj-89750	272	12	rangeland	rangeland	NOUN
esrj-89750	272	13	indicators	indicator	NOUN
esrj-89750	272	14	:	:	PUNCT
esrj-89750	272	15	application	application	NOUN
esrj-89750	272	16	of	of	ADP
esrj-89750	272	17	bayesian	bayesian	NOUN
esrj-89750	272	18	additive	additive	ADJ
esrj-89750	272	19	regression	regression	NOUN
esrj-89750	272	20	trees	tree	NOUN
esrj-89750	272	21	.	.	PUNCT
esrj-89750	273	1	rangeland	rangeland	NOUN
esrj-89750	273	2	ecology	ecology	NOUN
esrj-89750	273	3	&	&	CCONJ
esrj-89750	273	4	management	management	NOUN
esrj-89750	273	5	,	,	PUNCT
esrj-89750	273	6	70(5):s1550742417300222	70(5):s1550742417300222	NOUN
esrj-89750	273	7	.	.	PUNCT
esrj-89750	273	8	palagan	palagan	PROPN
esrj-89750	273	9	,	,	PUNCT
esrj-89750	273	10	c.	c.	PROPN
esrj-89750	273	11	a.	a.	PROPN
esrj-89750	273	12	,	,	PUNCT
esrj-89750	273	13	&	&	CCONJ
esrj-89750	273	14	geetha	geetha	PROPN
esrj-89750	273	15	,	,	PUNCT
esrj-89750	273	16	k.	k.	PROPN
esrj-89750	274	1	p.	p.	PROPN
esrj-89750	274	2	(	(	PUNCT
esrj-89750	274	3	2016	2016	NUM
esrj-89750	274	4	)	)	PUNCT
esrj-89750	274	5	a	a	DET
esrj-89750	274	6	prediction	prediction	NOUN
esrj-89750	274	7	method	method	NOUN
esrj-89750	274	8	using	use	VERB
esrj-89750	274	9	instantaneous	instantaneous	ADJ
esrj-89750	274	10	mixing	mixing	NOUN
esrj-89750	274	11	plus	plus	CCONJ
esrj-89750	274	12	auto	auto	NOUN
esrj-89750	274	13	regressive	regressive	ADJ
esrj-89750	274	14	approach	approach	NOUN
esrj-89750	274	15	in	in	ADP
esrj-89750	274	16	frequency	frequency	NOUN
esrj-89750	274	17	domain	domain	NOUN
esrj-89750	274	18	for	for	ADP
esrj-89750	274	19	separating	separate	VERB
esrj-89750	274	20	speech	speech	NOUN
esrj-89750	274	21	signals	signal	NOUN
esrj-89750	274	22	by	by	ADP
esrj-89750	274	23	short	short	ADJ
esrj-89750	274	24	time	time	NOUN
esrj-89750	274	25	fourier	fourier	X
esrj-89750	274	26	transform	transform	NOUN
esrj-89750	274	27	.	.	PUNCT
esrj-89750	275	1	biomedical	biomedical	ADJ
esrj-89750	275	2	research	research	PROPN
esrj-89750	275	3	-	-	PUNCT
esrj-89750	275	4	india	india	PROPN
esrj-89750	275	5	,	,	PUNCT
esrj-89750	275	6	27(4):1216	27(4):1216	NUM
esrj-89750	275	7	-	-	SYM
esrj-89750	275	8	1222	1222	NUM
esrj-89750	275	9	.	.	PUNCT
esrj-89750	276	1	pritikin	pritikin	PROPN
esrj-89750	276	2	,	,	PUNCT
esrj-89750	276	3	j.	j.	PROPN
esrj-89750	276	4	n.	n.	PROPN
esrj-89750	276	5	,	,	PUNCT
esrj-89750	276	6	brick	brick	NOUN
esrj-89750	276	7	,	,	PUNCT
esrj-89750	276	8	t.	t.	PROPN
esrj-89750	276	9	r.	r.	PROPN
esrj-89750	276	10	,	,	PUNCT
esrj-89750	276	11	&	&	CCONJ
esrj-89750	276	12	neale	neale	PROPN
esrj-89750	276	13	,	,	PUNCT
esrj-89750	276	14	m.	m.	NOUN
esrj-89750	276	15	c.	c.	PROPN
esrj-89750	276	16	(	(	PUNCT
esrj-89750	276	17	2018	2018	NUM
esrj-89750	276	18	)	)	PUNCT
esrj-89750	276	19	.	.	PUNCT
esrj-89750	277	1	multivariate	multivariate	VERB
esrj-89750	277	2	normal	normal	ADJ
esrj-89750	277	3	maximum	maximum	ADJ
esrj-89750	277	4	likelihood	likelihood	NOUN
esrj-89750	277	5	with	with	ADP
esrj-89750	277	6	both	both	CCONJ
esrj-89750	277	7	ordinal	ordinal	ADJ
esrj-89750	277	8	and	and	CCONJ
esrj-89750	277	9	continuous	continuous	ADJ
esrj-89750	277	10	variables	variable	NOUN
esrj-89750	277	11	,	,	PUNCT
esrj-89750	277	12	and	and	CCONJ
esrj-89750	277	13	data	datum	NOUN
esrj-89750	277	14	missing	miss	VERB
esrj-89750	277	15	at	at	ADP
esrj-89750	277	16	random	random	ADJ
esrj-89750	277	17	.	.	PUNCT
esrj-89750	278	1	behavior	behavior	NOUN
esrj-89750	278	2	research	research	NOUN
esrj-89750	278	3	methods	method	NOUN
esrj-89750	278	4	,	,	PUNCT
esrj-89750	278	5	50(2):490	50(2):490	PROPN
esrj-89750	278	6	-	-	PUNCT
esrj-89750	278	7	500	500	NUM
esrj-89750	278	8	.	.	PUNCT
esrj-89750	279	1	sato	sato	PROPN
esrj-89750	279	2	,	,	PUNCT
esrj-89750	279	3	a.	a.	NOUN
esrj-89750	279	4	(	(	PUNCT
esrj-89750	279	5	2012	2012	NUM
esrj-89750	279	6	)	)	PUNCT
esrj-89750	279	7	.	.	PUNCT
esrj-89750	280	1	a	a	DET
esrj-89750	280	2	method	method	NOUN
esrj-89750	280	3	to	to	PART
esrj-89750	280	4	quantify	quantify	VERB
esrj-89750	280	5	risks	risk	NOUN
esrj-89750	280	6	of	of	ADP
esrj-89750	280	7	financial	financial	ADJ
esrj-89750	280	8	assets	asset	NOUN
esrj-89750	280	9	:	:	PUNCT
esrj-89750	280	10	an	an	DET
esrj-89750	280	11	empirical	empirical	ADJ
esrj-89750	280	12	analysis	analysis	NOUN
esrj-89750	280	13	of	of	ADP
esrj-89750	280	14	japanese	japanese	ADJ
esrj-89750	280	15	security	security	NOUN
esrj-89750	280	16	prices	price	NOUN
esrj-89750	280	17	.	.	PUNCT
esrj-89750	281	1	2011	2011	NUM
esrj-89750	281	2	international	international	ADJ
esrj-89750	281	3	conference	conference	NOUN
esrj-89750	281	4	on	on	ADP
esrj-89750	281	5	management	management	NOUN
esrj-89750	281	6	,	,	PUNCT
esrj-89750	281	7	manufacturing	manufacturing	NOUN
esrj-89750	281	8	and	and	CCONJ
esrj-89750	281	9	materials	material	NOUN
esrj-89750	281	10	engineering	engineering	NOUN
esrj-89750	281	11	,	,	PUNCT
esrj-89750	281	12	icmmm	icmmm	ADJ
esrj-89750	281	13	2011	2011	NUM
esrj-89750	281	14	,	,	PUNCT
esrj-89750	281	15	december	december	PROPN
esrj-89750	281	16	8	8	NUM
esrj-89750	281	17	,	,	PUNCT
esrj-89750	281	18	2011	2011	NUM
esrj-89750	281	19	december	december	PROPN
esrj-89750	281	20	10	10	NUM
esrj-89750	281	21	,	,	PUNCT
esrj-89750	281	22	2011	2011	NUM
esrj-89750	281	23	,	,	PUNCT
esrj-89750	281	24	zhengzhou	zhengzhou	PROPN
esrj-89750	281	25	,	,	PUNCT
esrj-89750	281	26	china	china	PROPN
esrj-89750	281	27	.	.	PUNCT
esrj-89750	281	28	table	table	PROPN
esrj-89750	281	29	3	3	NUM
esrj-89750	281	30	.	.	PUNCT
esrj-89750	282	1	network	network	NOUN
esrj-89750	282	2	parameters	parameter	NOUN
esrj-89750	282	3	after	after	SCONJ
esrj-89750	282	4	training	train	VERB
esrj-89750	282	5	series	series	NOUN
esrj-89750	282	6	s	s	PART
esrj-89750	282	7	-	-	PUNCT
esrj-89750	282	8	surface	surface	NOUN
esrj-89750	282	9	number	number	NOUN
esrj-89750	282	10	included	include	VERB
esrj-89750	282	11	in	in	ADP
esrj-89750	282	12	s	s	NOUN
esrj-89750	282	13	-	-	NOUN
esrj-89750	282	14	layer	layer	NOUN
esrj-89750	282	15	s	s	NOUN
esrj-89750	282	16	-	-	PUNCT
esrj-89750	282	17	element	element	ADJ
esrj-89750	282	18	level	level	NOUN
esrj-89750	282	19	1	1	NUM
esrj-89750	282	20	16	16	NUM
esrj-89750	282	21	32	32	NUM
esrj-89750	282	22	level	level	NOUN
esrj-89750	282	23	2	2	NUM
esrj-89750	282	24	19	19	NUM
esrj-89750	282	25	290	290	NUM
esrj-89750	282	26	level	level	NOUN
esrj-89750	282	27	3	3	NUM
esrj-89750	282	28	15	15	NUM
esrj-89750	282	29	272	272	NUM
esrj-89750	282	30	level	level	NOUN
esrj-89750	282	31	4	4	NUM
esrj-89750	282	32	7	7	NUM
esrj-89750	282	33	89	89	NUM
esrj-89750	282	34	365maximum	365maximum	NUM
esrj-89750	282	35	likelihood	likelihood	NOUN
esrj-89750	282	36	classification	classification	NOUN
esrj-89750	282	37	of	of	ADP
esrj-89750	282	38	soil	soil	NOUN
esrj-89750	282	39	remote	remote	ADJ
esrj-89750	282	40	sensing	sense	VERB
esrj-89750	282	41	image	image	NOUN
esrj-89750	282	42	based	base	VERB
esrj-89750	282	43	on	on	ADP
esrj-89750	282	44	deep	deep	ADJ
esrj-89750	282	45	learning	learning	NOUN
esrj-89750	282	46	temmer	temmer	NOUN
esrj-89750	282	47	,	,	PUNCT
esrj-89750	282	48	m.	m.	NOUN
esrj-89750	282	49	,	,	PUNCT
esrj-89750	282	50	thalmann	thalmann	PROPN
esrj-89750	282	51	,	,	PUNCT
esrj-89750	282	52	j.	j.	PROPN
esrj-89750	282	53	k.	k.	PROPN
esrj-89750	282	54	,	,	PUNCT
esrj-89750	282	55	dissauer	dissauer	PROPN
esrj-89750	282	56	,	,	PUNCT
esrj-89750	282	57	k.	k.	PROPN
esrj-89750	282	58	,	,	PUNCT
esrj-89750	282	59	veronig	veronig	PROPN
esrj-89750	282	60	,	,	PUNCT
esrj-89750	282	61	a.	a.	NOUN
esrj-89750	282	62	m.	m.	NOUN
esrj-89750	282	63	,	,	PUNCT
esrj-89750	282	64	tschernitz	tschernitz	PROPN
esrj-89750	282	65	,	,	PUNCT
esrj-89750	282	66	j.	j.	PROPN
esrj-89750	282	67	,	,	PUNCT
esrj-89750	282	68	hinterreiter	hinterreiter	PROPN
esrj-89750	282	69	,	,	PUNCT
esrj-89750	282	70	j.	j.	PROPN
esrj-89750	282	71	,	,	PUNCT
esrj-89750	282	72	&	&	CCONJ
esrj-89750	282	73	rodriguez	rodriguez	PROPN
esrj-89750	282	74	,	,	PUNCT
esrj-89750	282	75	l.	l.	PROPN
esrj-89750	282	76	(	(	PUNCT
esrj-89750	282	77	2017	2017	NUM
esrj-89750	282	78	)	)	PUNCT
esrj-89750	282	79	.	.	PUNCT
esrj-89750	283	1	on	on	ADP
esrj-89750	283	2	flare	flare	NOUN
esrj-89750	283	3	-	-	PUNCT
esrj-89750	283	4	cme	cme	PROPN
esrj-89750	283	5	characteristics	characteristic	NOUN
esrj-89750	283	6	from	from	ADP
esrj-89750	283	7	sun	sun	NOUN
esrj-89750	283	8	to	to	ADP
esrj-89750	283	9	earth	earth	NOUN
esrj-89750	283	10	combining	combine	VERB
esrj-89750	283	11	remote	remote	ADV
esrj-89750	283	12	-	-	PUNCT
esrj-89750	283	13	sensing	sense	VERB
esrj-89750	283	14	image	image	NOUN
esrj-89750	283	15	data	datum	NOUN
esrj-89750	283	16	with	with	ADP
esrj-89750	283	17	in	in	ADP
esrj-89750	283	18	situ	situ	ADJ
esrj-89750	283	19	measurements	measurement	NOUN
esrj-89750	283	20	supported	support	VERB
esrj-89750	283	21	by	by	ADP
esrj-89750	283	22	modeling	modeling	NOUN
esrj-89750	283	23	.	.	PUNCT
esrj-89750	284	1	solar	solar	ADJ
esrj-89750	284	2	physics	physics	PROPN
esrj-89750	284	3	,	,	PUNCT
esrj-89750	284	4	292(7):93	292(7):93	NUM
esrj-89750	284	5	.	.	PUNCT
esrj-89750	285	1	wang	wang	PROPN
esrj-89750	285	2	,	,	PUNCT
esrj-89750	285	3	t.	t.	PROPN
esrj-89750	285	4	,	,	PUNCT
esrj-89750	285	5	strobl	strobl	NOUN
esrj-89750	285	6	,	,	PUNCT
esrj-89750	285	7	c.	c.	PROPN
esrj-89750	285	8	,	,	PUNCT
esrj-89750	285	9	zeileis	zeileis	PROPN
esrj-89750	285	10	,	,	PUNCT
esrj-89750	285	11	a.	a.	NOUN
esrj-89750	285	12	,	,	PUNCT
esrj-89750	285	13	&	&	CCONJ
esrj-89750	285	14	merkle	merkle	NOUN
esrj-89750	285	15	,	,	PUNCT
esrj-89750	285	16	e.	e.	PROPN
esrj-89750	285	17	c.	c.	PROPN
esrj-89750	285	18	(	(	PUNCT
esrj-89750	285	19	2017	2017	NUM
esrj-89750	285	20	)	)	PUNCT
esrj-89750	285	21	.	.	PUNCT
esrj-89750	286	1	score	score	NOUN
esrj-89750	286	2	-	-	PUNCT
esrj-89750	286	3	based	base	VERB
esrj-89750	286	4	tests	test	NOUN
esrj-89750	286	5	of	of	ADP
esrj-89750	286	6	differential	differential	ADJ
esrj-89750	286	7	item	item	NOUN
esrj-89750	286	8	functioning	function	VERB
esrj-89750	286	9	via	via	ADP
esrj-89750	286	10	pairwise	pairwise	NOUN
esrj-89750	286	11	maximum	maximum	ADJ
esrj-89750	286	12	likelihood	likelihood	NOUN
esrj-89750	286	13	estimation	estimation	NOUN
esrj-89750	286	14	.	.	PUNCT
esrj-89750	287	1	psychometrika	psychometrika	NOUN
esrj-89750	287	2	,	,	PUNCT
esrj-89750	287	3	83(1):132	83(1):132	NOUN
esrj-89750	287	4	-	-	SYM
esrj-89750	287	5	155	155	NUM
esrj-89750	287	6	.	.	PUNCT
esrj-89750	288	1	xu	xu	PROPN
esrj-89750	288	2	,	,	PUNCT
esrj-89750	288	3	y.	y.	PROPN
esrj-89750	288	4	,	,	PUNCT
esrj-89750	288	5	smith	smith	PROPN
esrj-89750	288	6	,	,	PUNCT
esrj-89750	288	7	s.	s.	PROPN
esrj-89750	288	8	e.	e.	PROPN
esrj-89750	288	9	,	,	PUNCT
esrj-89750	288	10	grunwald	grunwald	PROPN
esrj-89750	288	11	,	,	PUNCT
esrj-89750	288	12	s.	s.	PROPN
esrj-89750	288	13	,	,	PUNCT
esrj-89750	288	14	abd	abd	PROPN
esrj-89750	288	15	-	-	PUNCT
esrj-89750	288	16	elrahman	elrahman	PROPN
esrj-89750	288	17	,	,	PUNCT
esrj-89750	288	18	a.	a.	NOUN
esrj-89750	288	19	,	,	PUNCT
esrj-89750	288	20	&	&	CCONJ
esrj-89750	288	21	wani	wani	PROPN
esrj-89750	288	22	,	,	PUNCT
esrj-89750	289	1	s.	s.	PROPN
esrj-89750	289	2	p.	p.	PROPN
esrj-89750	289	3	(	(	PUNCT
esrj-89750	289	4	2017	2017	NUM
esrj-89750	289	5	)	)	PUNCT
esrj-89750	289	6	.	.	PUNCT
esrj-89750	290	1	evaluating	evaluate	VERB
esrj-89750	290	2	the	the	DET
esrj-89750	290	3	effect	effect	NOUN
esrj-89750	290	4	of	of	ADP
esrj-89750	290	5	remote	remote	ADJ
esrj-89750	290	6	sensing	sense	VERB
esrj-89750	290	7	image	image	NOUN
esrj-89750	290	8	spatial	spatial	ADJ
esrj-89750	290	9	resolution	resolution	NOUN
esrj-89750	290	10	on	on	ADP
esrj-89750	290	11	soil	soil	NOUN
esrj-89750	290	12	exchangeable	exchangeable	ADJ
esrj-89750	290	13	potassium	potassium	NOUN
esrj-89750	290	14	prediction	prediction	NOUN
esrj-89750	290	15	models	model	NOUN
esrj-89750	290	16	in	in	ADP
esrj-89750	290	17	smallholder	smallholder	NOUN
esrj-89750	290	18	farm	farm	NOUN
esrj-89750	290	19	settings	setting	NOUN
esrj-89750	290	20	.	.	PUNCT
esrj-89750	291	1	journal	journal	NOUN
esrj-89750	291	2	of	of	ADP
esrj-89750	291	3	environmental	environmental	ADJ
esrj-89750	291	4	management	management	NOUN
esrj-89750	291	5	,	,	PUNCT
esrj-89750	291	6	200:423	200:423	PROPN
esrj-89750	291	7	.	.	PUNCT
esrj-89750	292	1	zhao	zhao	PROPN
esrj-89750	292	2	,	,	PUNCT
esrj-89750	292	3	y.	y.	PROPN
esrj-89750	292	4	,	,	PUNCT
esrj-89750	292	5	zeng	zeng	PROPN
esrj-89750	292	6	,	,	PUNCT
esrj-89750	292	7	x.	x.	PROPN
esrj-89750	292	8	,	,	PUNCT
esrj-89750	292	9	qiang	qiang	PROPN
esrj-89750	292	10	,	,	PUNCT
esrj-89750	292	11	g.	g.	PROPN
esrj-89750	292	12	,	,	PUNCT
esrj-89750	292	13	&	&	CCONJ
esrj-89750	292	14	xu	xu	PROPN
esrj-89750	292	15	,	,	PUNCT
esrj-89750	292	16	m.	m.	NOUN
esrj-89750	292	17	(	(	PUNCT
esrj-89750	292	18	2018	2018	NUM
esrj-89750	292	19	)	)	PUNCT
esrj-89750	292	20	.	.	PUNCT
esrj-89750	293	1	an	an	DET
esrj-89750	293	2	integration	integration	NOUN
esrj-89750	293	3	of	of	ADP
esrj-89750	293	4	fast	fast	ADJ
esrj-89750	293	5	alignment	alignment	NOUN
esrj-89750	293	6	and	and	CCONJ
esrj-89750	293	7	maximum	maximum	ADJ
esrj-89750	293	8	-	-	PUNCT
esrj-89750	293	9	likelihood	likelihood	NOUN
esrj-89750	293	10	methods	method	NOUN
esrj-89750	293	11	for	for	ADP
esrj-89750	293	12	electron	electron	NOUN
esrj-89750	293	13	subtomogram	subtomogram	NOUN
esrj-89750	293	14	averaging	averaging	NOUN
esrj-89750	293	15	and	and	CCONJ
esrj-89750	293	16	classification	classification	NOUN
esrj-89750	293	17	.	.	PUNCT
esrj-89750	294	1	bioinformatics	bioinformatic	NOUN
esrj-89750	294	2	,	,	PUNCT
esrj-89750	294	3	34(13):i227	34(13):i227	NUM
esrj-89750	294	4	-	-	SYM
esrj-89750	294	5	i236	i236	PROPN
esrj-89750	294	6	.	.	PUNCT
esrj-89750	295	1	zhu	zhu	PROPN
esrj-89750	295	2	,	,	PUNCT
esrj-89750	295	3	d.	d.	PROPN
esrj-89750	295	4	(	(	PUNCT
esrj-89750	295	5	2016	2016	NUM
esrj-89750	295	6	)	)	PUNCT
esrj-89750	295	7	a	a	DET
esrj-89750	295	8	two	two	NUM
esrj-89750	295	9	-	-	PUNCT
esrj-89750	295	10	component	component	NOUN
esrj-89750	295	11	mixture	mixture	NOUN
esrj-89750	295	12	model	model	NOUN
esrj-89750	295	13	for	for	ADP
esrj-89750	295	14	density	density	NOUN
esrj-89750	295	15	estimation	estimation	NOUN
esrj-89750	295	16	and	and	CCONJ
esrj-89750	295	17	classification	classification	NOUN
esrj-89750	295	18	.	.	PUNCT
esrj-89750	296	1	journal	journal	NOUN
esrj-89750	296	2	of	of	ADP
esrj-89750	296	3	interdisciplinary	interdisciplinary	ADJ
esrj-89750	296	4	mathematics	mathematic	NOUN
esrj-89750	296	5	,	,	PUNCT
esrj-89750	296	6	19(2):311319	19(2):311319	NUM
esrj-89750	296	7	.	.	PUNCT
esrj-89750	297	1	zhu	zhu	PROPN
esrj-89750	297	2	,	,	PUNCT
esrj-89750	297	3	h.	h.	PROPN
esrj-89750	297	4	,	,	PUNCT
esrj-89750	297	5	jiao	jiao	PROPN
esrj-89750	297	6	,	,	PUNCT
esrj-89750	297	7	l.	l.	PROPN
esrj-89750	297	8	,	,	PUNCT
esrj-89750	297	9	ma	ma	PROPN
esrj-89750	297	10	,	,	PUNCT
esrj-89750	297	11	w.	w.	PROPN
esrj-89750	297	12	,	,	PUNCT
esrj-89750	297	13	liu	liu	PROPN
esrj-89750	297	14	,	,	PUNCT
esrj-89750	297	15	f.	f.	PROPN
esrj-89750	297	16	,	,	PUNCT
esrj-89750	297	17	&	&	CCONJ
esrj-89750	297	18	zhao	zhao	PROPN
esrj-89750	297	19	,	,	PUNCT
esrj-89750	297	20	w.	w.	PROPN
esrj-89750	297	21	(	(	PUNCT
esrj-89750	297	22	2019	2019	NUM
esrj-89750	297	23	)	)	PUNCT
esrj-89750	297	24	.	.	PUNCT
esrj-89750	298	1	a	a	DET
esrj-89750	298	2	novel	novel	ADJ
esrj-89750	298	3	neural	neural	ADJ
esrj-89750	298	4	network	network	NOUN
esrj-89750	298	5	for	for	ADP
esrj-89750	298	6	remote	remote	ADJ
esrj-89750	298	7	sensing	sense	VERB
esrj-89750	298	8	image	image	NOUN
esrj-89750	298	9	matching	matching	NOUN
esrj-89750	298	10	.	.	PUNCT
esrj-89750	299	1	ieee	ieee	NOUN
esrj-89750	299	2	transactions	transaction	NOUN
esrj-89750	299	3	on	on	ADP
esrj-89750	299	4	neural	neural	ADJ
esrj-89750	299	5	networks	network	NOUN
esrj-89750	299	6	and	and	CCONJ
esrj-89750	299	7	learning	learning	NOUN
esrj-89750	299	8	systems	system	NOUN
esrj-89750	299	9	,	,	PUNCT
esrj-89750	299	10	pp(99):1	pp(99):1	NOUN
esrj-89750	299	11	-	-	SYM
esrj-89750	299	12	13	13	NUM
esrj-89750	299	13	.	.	PUNCT