Начиная с начала 2000 года осуществляется внедрение GHIS в здравоохранении, в рамках принятого проекта о реформирование информ


Mathematical Problems of Computer Science 39, 125--128, 2013. 

 

125 

 

 

 

 

 

Moving Objects Detection on Colored Video Images 
 

Suren B. Alaverdyan 

 

Institute for Informatics and Automation Problems of NAS RA 

e-mail: souren@ipia.sci.am 
 

Abstract 

 
The  background covering formulas of images are obtained. With 

their help the algorithm of detection of  moving objects on colored video 

images is given. 

Keywords: image, detection.             
 

 

 

1. Introduction 
 

 When solving various problems of colored image processing, the image is represented in a 

gray scale   (     )  ⁄  or                   to accelerate processing and reduce 
the used memory excluding the  possible damage of the original image (loss of contour points or 

appearance of false contour points). 

 In Fig. 2 the difference between the original and the gray scale images is given, where the 

values of  (   )-th point are as follows: 
 

    
  

|               |

 
     

  
|               |

 
     

  
|               |

 
   

 

 
                                                   Fig 1. Original             Fig. 2. Difference  

 

It is clear that the components artificially added to the original image, influence the result of the 

problem being solved. . 

 While obtaining the image in image processing theory, the image device accuracy, image 

quantization and, certainly, the environment which influences the  formation of the image noise, 

are principally important. , Moreover, since for the problems the detection of objects on images 

or their segmentation the problem of general filters' synthesis is not solved, then  in practice to 

mailto:souren@ipia.sci.am


Moving Objects Detection on Colored Video Images 

 
126 

solve these problems different approaches are used - -threshold restriction, brightness interval 

transformation, cluster analysis etc. – [1, 2].    

 In this article a new approach of moving objects detection on video images, obtained by a 

stationary camera, is given. 

 

2. Image Background Change by Standard Image 
 

  A mathematical definition of the background does not exist, it is interpreted as the 

background of image [3, 4]. 

  Let   {          ̅̅ ̅̅ ̅      ̅̅̅̅̅}                                     (              ), (red, 

green and blue colour channels of a point), and      the standard image with the size     . To 
reduce the background of the image           background of the image     we find the following 
statistical characteristics of both images: 

mean values 

   
 

  
∑ ∑     

 
   

 
    ,       

 

  
∑ ∑     

 
   

 
   , 

mean square deviations  

   √
 

  
∑ (      )

 
   ,      √

 

  
∑ (      )

 
   . 

The general formula for the new values of the image            background of the image 
     represented as 
 

                                     
  

  

  
(       )    ,         ̅̅̅̅ ̅̅       ̅̅̅̅ ̅ ,                                                

(1)        
 

let       
    if       

     and let       
        if       

       

Below we give an example of the background change by formula (1). 

 

 
      Fig. 3. Original image                     Fig. 4. Standard image                 Fig. 5. Resulting image 

 

 Thus, the parameters (  ,   ) can determine the approximate value of the image 
background.  Considering statistical moments of higher orders one can obtain a more exact value 

of background.   

 

3. Object Detecting 
 

To avoid artificially covering the image with new data (Fig. 2) we do not use in this work a 

gray scale image to detect objects. 

There are two principal methods used to detect moving video objects on video streams: 

determination of image stationary part and statistical methods with threshold restriction. 

In this work to detect the objects the original image is transformed into a negative image 

background space by formula (1) to reduce the probability of loss of dark colored objects. Since 

the negative mean value is equal  to         the formula (1)  takes the following form: 



S. Alaverdyan 

 
127 

                             
  

  

        
(       )        ,      ̅̅̅̅ ̅̅       ̅̅̅̅ ̅ .              

 

To obtain a binary (black-white) image, calculate the inter frame gradient using the following 

formula: 

       {|    
( )

  
   
(   )

|    
   
(     )

} ,         ̅̅ ̅̅ ̅̅       ̅̅̅̅ ̅  

 

 

where 

  
   
(     )

 
 

 
√∑ ∑ (   

( )
    

(   )
)
 

   
     

   
      ,   

 

  - is the frame number, (   ) interval of  (   ) point is used. The threshold is determined as 

follows:      
∑ ∑     

 
   

 
   

  
  or        . Both of threshold values give good results. 

Binary image    {          ̅̅̅̅ ̅̅       ̅̅̅̅ ̅} is obtained as follows: 
 

     {
            

               
 

 

Below the results of object detecting algorithms are given. 

 

 
                     Fig. 6. Original image with negative background                                                          Fig. 7. Objects       

 

 
                     Fig. 8. Original image with negative background                                                          Fig. 9.  Objects       

 

 

 



Moving Objects Detection on Colored Video Images 

 
128 

References 
 

[1]  У. Прэтт, Цифровая обработка изображений, М. Мир, т.2, 1982. 

[2] T. Ko, S. Soatto and D. Estrin, “Background subtraction on distributions”, Proceedings of the 

10th European Conference on Computer Vision: Part III, pp. 276-289, 2008. 

[3] V. Konushin and A. Konushin, “Improvement of background subtraction by mask   

constraints”, Proc. GraphiCon,  pp. 96-99, 2010. 

[4]  P Viola, M Jones, “Rapid object detection using a boosted cascade of simple features”, IEEE 

Computer Society Conference on Computer Vision and Pattern Recognition, pp. 511-518, 

2001. 

 

       

Submitted 30.11.2012, accepted 15.02.2013. 

 

 
 

Տեսաշարում գունավոր պատկերների վրա շարժվող 

օբյեկտների հայտնաբերում 
 

Ս.  Ալավերդյան 

 
Ամփոփում 

 

Աշխատանքում ստացված են պատկերների ֆոնային տեղափոխության բանաձևեր,  

բերված է նրանց կիրառության միջոցով տեսաշարում գունավոր պատկերների վրա 

շարժվող օբյեկտների հայտնաբերման ալգորիթմը: 
 

 

 

 

Обнаружение движущихся объектов на  

видеоизображениях 
 

С.  Алавердян 

 

Аннотация 
 

Приведен новый алгоритм нахождения движущихся объектов на цвeтных 

видеоизображениях. Приведены также результаты работы алгоритма.