Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 6s (2024) 427 https://internationalpubls.com An Improved Mathematical Model by Applying Machine Learning Algorithms for Identifying Various Medicinal Plants and Raw Materials Dr. Srinivasa Babu Kasturi1, Sreedhar Burada2, Dr. Sowmyashree M.S3, Sharath.S4, Dr. M. Sunil Kumar5*,Dr. D. Ganesh5 1Professor, Department of Computer Science and Engineering, Nalla Narasimha Reddy Education Society’s Group of Institutions, Kasturisrinu.babu@gmail.com 2Assistant Professor, Dept. of CSE, Koneru Lakshmaiah Education Foundation, Vaddeswaram, Guntur Dist., Andhra Pradesh,India. sreedharburada1@gmail.com 3Assistant Professor, Department of Electronics and Communication, GEC Chamarajanagar VTU,Belgaum, KA, India, sowmya.mtech@gmail.com 4Assistant Professor, Department of Electronics and Communication, GEC Chamarajanagar VTU,Belgaum, KA, India. ss.sharath@gmail.com 5*Professor & Programme Head, Department of CSE, School of Computing, Mohan Babu University (erstwhile Sree Vidyanikethan Engineering College),Tirupathi, AP, India. sunilmalchi1@gmail.com 5Associate Professor, Department of CSE, School of Computing, Mohan Babu University (erstwhile Sree Vidyanikethan Engineering College),Tirupathi, AP, India. dgani05@gmail.com Article History: Received: 19-06-2024 Revised: 26-07-2024 Accepted: 05-08-2024 Abstract: The utilization of medicinal plants holds significant promise for healthcare and pharmaceutical applications. However, the accurate identification of diverse plant species or raw materials can be challenging, necessitating innovative solutions. This project presents a novel approach for the Recognizable proof of Therapeutic Plants or Crude Materials through Picture Preparing, utilizing progressed Machine Learning Calculations. with a wealthy legacy of flower differing qualities, is well-known for its medicinal plant resources, but distinguishing evidence is one of the most challenging challenges in Ayurvedic Pharmaceutics. Some unrefined drugs are sold under the same name on the marketplace, causing confusion and misidentification. Indeed, collectors and dealers are not entirely aware of the distinct morphological appearance or differentiating qualities of the numerous substances due to their frequent and geographical accessibility, as well as comparable properties. In this technique, a computer capable of differentiating diverse medicinal plants/raw materials using Image Preparation Using Distinctive Machine Learning Calculations would be quite beneficial. It will be helpful at every stage of the supply chain of crude cloth being used within the framework. Herre the picture preparing strategy is considered as the most strategy for classifying diverse plants of diverse characteristics or particular parcels or locales of the plant clears out which can be at that point recognized through picture handling. Distinguishing proof of plants through plant takes off on the premise of their shape, colour, and surface highlights utilizing computerized picture handling methods for classifying the plant species utilizing diverse machine learning calculations. This investigates work incorporates checking on numerous pictures preparing strategies to utilize machine learning to recognize different plants utilizing it take off highlight within the shape of a picture... Keywords: Medicinal plants identification, Image processing, Machine learning algorithms, Colour features, Pattern recognition, Leaf texture analysis, Computer vision. mailto:sowmya.mtech@gmail.com Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 6s (2024) 428 https://internationalpubls.com 1. Introduction For eras, therapeutic plants have been a fundamental portion of conventional medication, advertising a wide extend of characteristic cures to individuals. To fully utilise these plants' medicinal potential, identification is a crucial first step. The need for precise and trustworthy ways of identifying medicinal plants grows as cultures move more and more towards holistic and sustainable healthcare practices. Identification of medicinal plants entails identifying and differentiating between different plant species according to their morphological, anatomical, chemical, and genetic traits. This procedure is necessary to guarantee the proper use of these plants in both conventional treatment and the creation of contemporary medications. Indigenous knowledge has been essential in many places of the world for the identification and use of therapeutic plants. But as traditional medicine gains popularity around the world, there is an increasing demand for methodical, scientific methods of plant identification. Identification of medicinal plants is hampered and facilitated by their high biodiversity. Similar physical characteristics are shared by many species, although differences can also exist within a species. For the purpose of successfully identifying medicinal plants, a multidisciplinary approach that combines conventional wisdom with contemporary scientific approaches is necessary. For the correct consider and administration of biodiversity, plant species recognizable proof is basic [21][23]. Diverse plant characteristics are surveyed as recognizable proof keys in a manual recognizable proof strategy, and these keys are at that point examined dynamically and adaptively to recognize distinctive plant species. All things considered, the manual method takes a long time and is repetitive. More effective ways to satisfy the needs of species identification have recently become necessary due to technological advancements. Examples of these methods include the creation of digital image processing and pattern recognition algorithms. Accurately automating plant species identification remains a challenge despite numerous studies already conducted. This introduction sets the stage for exploring the diverse methodologies employed within the recognizable proof of therapeutic plants, highlighting the significance of bridging conventional information with modern logical progressions in this vital field[22][24]. Therapeutic plants are quickly biting the dust off, a few of which are indeed close termination Executing solid preservation measures for therapeutic plants has ended up a pressing need [25]. Nature has given us restorative plants which have played a noteworthy part for thousands of a long time within the battle against diseases. However, distinguishing plants within the field isn't a errand that everybody can perform, not indeed college understudies with important majors. As it were youthful and middle-aged specialists with a long time of pertinent and ceaseless work encounter within the field can be competent for that errand. therapeutic plants are nothing but plants of restorative values, it can be expectedly concluded that the strategy is additionally viable in division of leaf pictures of all plants essential therapeutic care. According to estimates, plant medications account for 80% of total drugs in India and China, but just 25% in the United States. The information about inborn plants is widely known by a few professionals, nation and local persons, and it just has to be proven[26]. Botanists use different plant traits as identifying proof keys in a manual identifiable proof preparation, which are evaluated sequentially and adaptively to distinguish plant species. (For Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 6s (2024) 429 https://internationalpubls.com example, the form, color, and number of petals). In any event, determining plant species from field perception necessitates extensive botanical knowledge, which is beyond the grasp of most nature lovers [27].As of late, taxonomists begun looking for more proficient strategies to meet species recognizable proof prerequisites, such as creating computerized picture handling and design acknowledgment strategies. Advanced picture preparing alludes to the utilize of calculations and strategies for operations such as picture upgrade, picture compression, picture investigation, mapping, and geo-referencing. Image-based strategies are considered a promising approach for species distinguishing proof. A client can take a picture of a plant within the field with the build-in camera of a portable gadget and examine it with an introduced acknowledgment application to recognize the species or at slightest to get a list of conceivable species in the event that a single coordinate is inconceivable. For case, leaf shape is exceptionally vital not as it were for species distinguishing proof, but moreover in other ponders, such as plant environment and physiology. Shape examination of blooms has moreover been considered for species distinguishing proof. Our proposed system (SVM) gives accurate results when compared to CNN [28]. The following are the uses of SVM: Image Classification: Support Vector Machines (SVM) are extensively employed in image classification applications, including handwritten digit identification, facial expression analysis, and object recognition. Text and Document Classification: Natural language processing activities such as spam detection, sentiment analysis, and text categorization are handled by Support Vector Machines (SVM). Bioinformatics: SVM is used in bioinformatics to predict protein structures, classify gene expression, and classify diseases using genetic data. Medical Diagnosis: SVM is used to diagnose diseases in the medical industry. For example, it is used to classify cancers based on patient data or medical imaging. 2.Literature Survey: Within the term paper [1] "Recognizable proof of Restorative Plant utilizing Picture Handling," the utilize of convolutional neural systems (CNN) for mechanized therapeutic plant identification is inspected. The consider, which was distributed within the Worldwide Diary of Progressed Inquire about in Science, Communication, and Innovation, highlights how pivotal it is to utilize machine learning procedures in arrange to urge past the disadvantages of manual distinguishing proof. Preprocessing, include extraction, demonstrate development, assessment, and arrangement are all portion of the recommended calculation. The CNN engineering, equipment and computer program prerequisites, and a list of pertinent academic distributions are too secured in this think about. The by and large objective of the venture is to make strides therapeutic plant distinguishing proof productivity and exactness through imaginative mechanical arrangements [29][30]. The document discusses the development [2] of an automated system for identifying and categorizing medicinal plants using the MCMLGP feature descriptor. When it comes to categorizing medicinal plants, the proposed approach—based on multi-channel complementary local gradient patterns—consistently outperforms existing descriptors. The BU Medicinal Plants dataset has ten Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 6s (2024) 430 https://internationalpubls.com unique plant species. The study demonstrates the effectiveness of MCMLGP in gathering colour information over many channels by using SVM classifiers with varying kernels. With 40 training photos, the suggested method achieves a recognition rate of 96.11%, indicating potential for automated medical plant identification systems [31][32]. Here, [3] the author referred to research of image processing on plant classification, and leaf recognition. Techniques include shape-texture analysis, neural network-based algorithms, and near- infrared spectroscopy for traditional Chinese medicinal items. Studies on leaf vein extraction, colour- edge characteristics for Indian medicinal plants, and a suggested technique for medicinal leaf identification with image processing algorithms are all included in the publication. The references provide a wide range of methods for recognizing and categorizing medicinal leaves since they address different facets of plant classification and image analysis. The study [4] looks into using dry leaf image analysis to identify medicinal plants and suggests a feature set combination. With a maximum recognition rate of 94.5%, the suggested approach makes use of 14 classifiers, with the Multi-Layer Perceptron (MLP) Neural Network demonstrating the highest accuracy. When evaluating green leaves, the front and back aspects are included to improve identification. The authors stress the need for more samples to increase classifier accuracy and propose future work incorporating a large medicinal plant database. Little or leafless plants, however, might not be able to use this strategy. Financial support and database assistance are acknowledged. The article [5] highlights the numerous conventional restorative applications of 42 species of restorative plants from the Yercaud slopes in Tamil Nadu, India. As per the life frame investigation, climbers are the foremost prevalent and clears out are the foremost utilized plant component. These plants' helpful esteem is emphasized, as they treat a run of sicknesses in rustic zones. The [6] use of log-Gabor filters in the analysis of leaf images of medicinal plants is examined in the cited paper. It explores a number of research projects and techniques that use these filters to extract features and segment textures. The study covers plant categorization techniques and includes datasets such as the Ghanaian Leaf Dataset. The use of homomorphic filters and contrast-limited adaptive histogram equalization in hybrid picture enhancing techniques is investigated. The study uses Random forests and support vector machines are two examples of machine learning techniques that to classify medicinal plants according to their leaf characteristics. The ponder highlights how significant machine learning and picture handling strategies are to the consider of restorative plants. The [7] study's main objective is to categorize 45 kinds of medicinal plants using deep learning methods and an Exception network that has already been trained. After 50 epochs, the model produced acceptable training (98.40%) and validation (97.65%) accuracies using 4682 leaf images. Predictions on random images, a confusion matrix, and loss and accuracy graphs are some of the outcomes. Potential future research on customized deep learning models is suggested by the paper. The references show a variety of methodologies used in different research investigate for the classification of therapeutic plants, counting convolutional neural systems, picture handling, and machine learning. The paper [8] presents a Convolutional Neural Network (CNN) method that uses TensorFlow to detect uncommon medicinal plants with 90% test accuracy. In order to highlight the importance of Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 6s (2024) 431 https://internationalpubls.com training epochs, the Department of Ayurveda offered a dataset that was constructed from plant photographs. Works on recognition systems, floral categorization, and neural network-based leaf identification are cited. The piece highlights the importance of proper plant identification in Ayurvedic medicine, and CNN's promise in this sector. The study [9] paper talks about a smartphone app called Medic Plant that uses artificial intelligence and computer vision to Identify the medicinal herbs. The application aims to improve traditional medicine by assisting in Identification of medicinal plants in Mauritius. The study contributes to sustainable development goals by highlighting the role of medicinal plants in global health. The authors study technology solutions for healthcare and biodiversity conservation using machine learning techniques such as convolutional neural networks for plant recognition. The Tertiary Education Commission is funding this work, which highlights the possibilities of combining technology and conventional medical knowledge. In this [10] Based on leaf characteristics, the project investigates machine learning applications for medicinal and aromatic plant species classification. In comparison to CART, KNN, and PNN, the Naive Bayes Classifier (NBC) had the highest accuracy (98.39% training, 98.00% testing) while employing a k=10 cross-validation technique. Gray, form, and fractal features all made substantial contributions, as shown by feature analysis. Although morphologically comparable leaves are difficult to distinguish, the work presents a novel feature extraction technique. By combining features, the suggested method improves the accuracy of plant recognition. The results support the efficiency of NBC and provide guidance for enhancing automated plant categorization systems. Table1: A Summary of Literature Research on Medicinal/Medicinal Plant Species Identification Using Image Processing Using Machine Learning Algorithms. Sl. No Method Dataset Remarks Ref. No 1 CNN Image.net The key is to analyse the effects of "bad images" on the proposed starting network, such as chaotic or simpler images. We also collect another suggested method's performance will be tested using a fuzzy tree picture. [6] 2 SVM Folio data set Our architecture is used to train images of plant leaves. This method may be necessary to improve training time and performance accuracy. [10] Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 6s (2024) 432 https://internationalpubls.com 3 CNN Kaggle Convolutional neural networks (CNNs) have garnered significant attention in only radiology as a result of their astounding accomplishments in a number of fields, including medical research. [15] 4 CNN laboratory data set (Colour Histogram-LBP, Colour Histogram HOG- PCA) We use the data to train the machine by extracting important features and classifications to identify medicinal plants, which affects the accuracy of the system. In this study, we analyse image processing methods for classifying and identifying plants. [18] 5 SVM Folio data sets (histogram of oriented gradients and local binary patterns) To train the plant leaf pictures using this architecture and investigate the impact of the parallel CNN design. To increase training times and accuracy performance, this strategy can be required. [19] 6 Deep learning (CNN) Wikipedia datasets Classifying plant diseases using digital images is an extremely challenging task. Deep learning methods, notably CNNs, appear to be capable of efficiently tackling the vast majority of the technical challenges associated with plant disease classification. [20] 3. Problem statement: Developing an automated system for Medicinal Plants' identification via image processing and machine learning. Aim is to streamline raw material authentication, enhance biodiversity conservation, and ensure herbal product quality. Challenges include algorithm robustness, real-time processing, and user-friendly interfaces for seamless integration into pharmaceutical and healthcare workflows. Mathematical statement is as follows: F:M∪R M= set of medicinal plant images R =set of corresponding raw material images The goal is to design a function F:M∪R→Classes that maps each image to its respective medicinal plant or raw material class using machine learning algorithms, optimizing for accuracy. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 6s (2024) 433 https://internationalpubls.com 4. Methodology: The methodology the identification of medicinal plants and raw materials is primarily composed of two steps: image processing and machine learning algorithms. The dataset, which includes images of plant leaves, is first pre-processed in order to extract pertinent characteristics including colour, shape, and texture. Figure1: Flow chart of Identifying the Medicinal plants using Machine learning Algorithms 4.1 Input Image: ➢ Gather top-notch photos of various raw materials and medicinal plants in regulated and varied settings (angles, lighting, and backgrounds). ➢ Provide precise species labels and possibly targeted regions of interest (such as leaves or flowers) when annotating photos. ➢ By using techniques like rotations, flips, noise addition, and colour tweaks, you can increase the size and diversity of your dataset. 4.2 Preprocessing and Feature Extraction: ➢ Resize, normalize, and format images (into a certain colour space or grayscale). ➢ Using methods such as thresholding, edge detection, or object detection models, isolate areas of interest. ▪ Shape: Perimeter, area, moments, Fourier descriptors. ▪ Colour: Mean, standard deviation, histograms of channels. ▪ Texture: Hara lick features, Gabor filters, LBP, deep learning features. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 6s (2024) 434 https://internationalpubls.com 4.3 Training Data set: ➢ Gather a diverse set of images for each medicinal plant species. Use a high-resolution camera to capture images of different parts of the plants. ➢ Includes leaves, flowers, and stems. Make sure the photographs include different lighting situations, backdrops, and viewpoints. 4.4 Testing and Evaluation: ➢ Evaluate the model's capacity to recognize unfamiliar plants or materials using a test set that was not used for training. ➢ To evaluate performance, compute recall, accuracy, precision, F1-score, and other pertinent metrics. ➢ Examine frequent mistakes to find areas where preprocessing, feature extraction, or model selection should be strengthened. 4.5 Classification: ➢ Based on the properties of the data and the intended outcomes, choose an appropriate machine learning algorithm (SVM). ➢ Using the labelled training data, train the model and adjust the hyperparameters to achieve optimal results. ➢ Evaluate the model's performance on a different validation set and tweak hyperparameters if necessary. 5. Data Collection: Here, we collect the samples, document a customary wisdom, and conduct field research to methodically gather information for medicinal plant identification. Take detailed notes, scan and photograph specimens for the herbarium. Collaborate with experts, manage information effectively, and ensure that moral principles are followed at all times while respecting local laws and traditions. Gather a variety of information about medicinal plants, such as their identification, distribution, historical applications, chemical makeup, cultivation, dangers, and conservation initiatives. The Mendeley dataset can be used to arrange the literature on medicinal plants, making identification easier and gathering relevant data. 6. Experimental Results: This study was conducted using MATLAB R2017b. A 400 x 600 jpeg formatted. A medicinal plant leaf library of 760 leaf photos from 30 distinct plant groupings is created. The colour, shape, and texture properties are calculated to provide the feature vector required to train the dataset. The following Tables show the accuracy based on different attributes for the input data set after it has been examined using a variety of SVM Kernel functions. The accuracy shows how many therapeutic plants were appropriately categorized. Formal attribute When shape features are taken into account alone, the classification accuracy is quite low. As low as 0% is the accuracy of the linear SVM. The Gaussian kernel Both quadratic and cubic SVM get the same 8% accuracy.20% accuracy is given by Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 6s (2024) 435 https://internationalpubls.com the Fine, Medium, and Coarse SVM kernel functions. The results of shape feature accuracy displayed in Table II. Table II: Shape Feature Accuracy SVM Kernel Functions Shape Feature Accuracy Linear SVM 6.0% Quadratic SVM 8.0% Cubic SVM 8.0% Fine Gaussian SVM Scored 20.0% Medium Gaussian SVM Scored 20.0% Coarse Gaussian SVM Scored 20.0% The Texture Features specify the SVM kernel functions. The accuracy yielded by linear, quadratic, and cubic SVM is 16.7%, whereas the accuracy yielded by fine, medium, and coarse SVM kernel functions is 83.3%. The results of Texture feature accuracy displayed in Table III. Table III: Texture Feature Accuracy SVM Kernel Functions Texture Feature Accuracy Linear SVM 16.7% Quadratic SVM 16.7% Cubic SVM Scored 16.7% Fine Gaussian SVM Scored 83.3% Medium Gaussian SVM Scored 83.3% Coarse Gaussian SVM Scored 83.3% While we are aware that various species differ in terms of leaf form, and as a plant grows, various species vary in terms of leaf size. This is considered to be the crucial point, which is the measurement of each leaf's length and width as mentioned in the section above. This provides the information needed to determine the leaf's Aspect ratio, centroid, area, perimeter, and circularity. Each feature has a unique attribute that affects the group's leaf structure. We A feature set was developed for five different neem leaf diameters. Because not all leaves are perfectly round or elliptical, the Roundness value is utilized to distinguish them. This displays the pixel's roundness, which was explained in the preceding section. Because some plants have broad leaves while others have narrow ones, they differ depending on the species. This is one of the characteristics that differentiate a leaf. Tulsi, peppermint, Bael, lemon balm, catnip, and stevia leaves were classified with accuracy rates of 99.10%, 99.80%, 98.40%, 99.90%, 99.40%, and 99.20%, respectively. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 6s (2024) 436 https://internationalpubls.com Figure1: This Bar graph showing the Accuracy for the medicinal leaves. We began our experiments using ROOs that were 220 × 220 in size. After that, in order to improve accuracy, we progressively raised the dimensions of ROOs. Finally, we observe the promising accuracy at ROOs 280 × 280, which covers the maximum quantity of critical information. Speckle noise was diminishing accuracy as ROO sizes increased. Finally, as shown in the image, the comparison analysis is performed to categorize the medicinal plant leaves have ROO sizes of 220 x 220 and 280 x 280. A comparative study of medicinal plant leaf classification using machine learning-based classifiers with ROO sizes of (220 x 220) and (280 × 280). Figure2: Comparative study of medicinal leaf categorization using machine learning-based classifiers. Six different the fused (multispectral + texture) feature collection includes medicinal plants Examples include Tulsi, peppermint, Bael, lemon balm, catnip, and stevia. The extensive laboratory setting produced a highly standardized and polished dataset. The recommended SVM approach can detect 96.66% of the leaf properties of medicinal plants, including form, colour, and texture. 7. Conclusion: First, we may collect additional photos each and every plant on the existing plant list before creating a more accurate recognition algorithm. Second, we could broaden the dataset to include over 100 unique medicinal plants, and third, we could provide the user with more information on each plant. As a result, a straightforward and reliable method for recognizing herbs is necessary. Herbal identification is expected to benefit from both computational and statistical methods. This non- destructive technique will be preferable for swiftly identifying herbs, particularly by those who cannot afford expensive analytical equipment. Medic Plant is the first app for identifying medicinal plants that does not need an internet connection to function. The majority of fieldwork takes place in high hills, woodlands, and distant areas where the internet is usually unavailable. We also employ the most classes while doing medicinal plant research, as opposed to many other studies that require Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 6s (2024) 437 https://internationalpubls.com specialized equipment to photograph leaves in certain situations. Furthermore, the suggested machine vision system has great promise for future usage in categorizing and identifying a wide range of agricultural goods based on key texture, shape, and colour features. References: [1]. U. Habiba, M. Howlader, M. Islam, R. Faisal, and M. Rahman, ‘Automatic Medicinal Plants Classification using Multi-channel Modified Local Gradient Pattern with SVM Classifier’, in 2019 Joint 8th International Conference on Informatics, Electronics & Vision (ICIEV) and 2019 3rd International Conference on Imaging, Vision & Pattern Recognition (icIVPR), 2019, pp. 6– 11. [2] P. Barik, S. Zatale, G. Kale, K. Devi, and P. Koli, ‘Identification of Medicinal Plant using Image Processing’, International Journal of Advanced Research in Science, Communication and Technology, pp. 93–96, 12 2022. [3]. P. M. Kumar, C. M. Surya, and V. P. Gopi, ‘Identification of ayurvedic medicinal plants by image processing of leaf samples’, in 2017 Third International Conference on Research in Computational Intelligence and Communication Networks (ICRCICN), 2017, pp. 231–238. [4]. F. Twum, Y. M. Missah, S. O. Oppong, and N. Ussiph, ‘Textural Analysis for Medicinal Plants Identification Using Log Gabor Filters’, IEEE Access, vol. 10, pp. 83204–83220, 2022. [5]. T. Vijayashree and A. Gopal, ‘Leaf identification for the extraction of medicinal qualities using image processing algorithm’, in 2017 International Conference on Intelligent Computing and Control (I2C2), 2017, pp. 1–4. [6]. R. Rao, M. Lahari, K. Sri, K. Srujana, and D. Yaswanth, ‘Identification of Medicinal Plants using Deep Learning’, International Journal for Research in Applied Science and Engineering Technology, vol. 10, pp. 306–322, 04 2022. [7]. Vishnu Kumar S, Jemima DK, Dharani S, Narayanan N, Thejan SM, Floristic analysis of Medicinal Plants from the Sholas of Nilgiris, Tamil Nadu, India, Journal of Drug Delivery and Therapeutics. 2022; 12(6):16-21 [8]. N. Bhelkar and A. Sharma, ‘Identification and classification of medicinal plants using leaf with deep convolutional neural networks’, international journal of health sciences, pp. 11596–11605, 10 2022. [9]. A. D. A. D. S. Jayalath, T. G. A. G. D. Amarawanshaline, D. P. Nawinna, P. V. D. Nadeeshan, and H. P. Jayasuriya, ‘Identification of Medicinal Plants by Visual Characteristics of Leaves and Flowers’, in 2019 14th Conference on Industrial and Information Systems (ICIIS), 2019, pp. 125–129. [10]. S. Pudaruth, F. Mahomoodally, N. Kissoon, and F. Chady, ‘Medic Plant: A mobile application for the recognition of medicinal plants from the Republic of Mauritius using deep learning in real-time’, IAES International Journal of Artificial Intelligence (IJ-AI), vol. 10, p. 938, 12 2021. [11]. G. Kayhan and E. Ergün, ‘Medicinal and Aromatic Plants Identification Using Machine Learning Methods’, Balkan Journal of Electrical and Computer Engineering, vol. 8, 01 2020. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 6s (2024) 438 https://internationalpubls.com [12]. R. Azadnia and K. Kheiralipour, ‘Recognition of leaves of different medicinal plant species using a robust image processing algorithm and artificial neural networks classifier’, Journal of Applied Research on Medicinal and Aromatic Plants, vol. 25, p. 100327, 2021. [13]. C. Amuthalingeswaran, M. Sivakumar, P. Renuga, S. Alexpandi, J. Elamathi, and S. S. Hari, ‘Identification of Medicinal Plant’s and Their Usage by Using Deep Learning’, in 2019 3rd International Conference on Trends in Electronics and Informatics (ICOEI), 2019, pp. 886–890. [14]. Mukherjee, G., Tudu, B. & Chatterjee, A. A convolutional neural network-driven computer vision system toward identification of species and maturity stage of medicinal leaves: case studies with Neem, Tulsi and Kalmegh leaves. Soft Computer 25, 14119–14138 (2021). [15]. D. Pechebovicz et al., ‘Plants recognition using embedded Convolutional Neural Networks on Mobile devices’, in 2020 IEEE International Conference on Industrial Technology (ICIT), 2020, pp. 674–679. [16]. R. Janani and A. Gopal, ‘Identification of selected medicinal plant leaves using image features and ANN’, in 2013 International Conference on Advanced Electronic Systems (ICAES), 2013, pp. 238–242. [17].Russel, N.S., Selvaraj, A. Leaf species and disease classification using multiscale parallel deep CNN architecture. Neural Computer & Applic 34, 19217–19237 (2022). [18]. J. G. Arnal Barbedo, ‘Plant disease identification from individual lesions and spots using deep learning’, Biosystems Engineering, vol. 180, pp. 96–107, 2019. [19]. C. Sivaranjani, L. Kalinathan, R. Amutha, R. S. Kathavarayan, and K. J. Jegadish Kumar, ‘Real-Time Identification of Medicinal Plants using Machine Learning Techniques’, in 2019 International Conference on Computational Intelligence in Data Science (ICCIDS), 2019, pp. 1– 4. [20]. R. Azadnia, A. Jahanbakhshi, S. Rashidi, M. Khajehzadeh, and P. Bazyar, ‘Developing an automated monitoring system for fast and accurate prediction of soil texture using an image- based deep learning network and machine vision system’, Measurement, vol. 190, p. 110669, 2022. [21] Davanam, G., Pavan Kumar, T., & Sunil Kumar, M. (2021). Novel Defense Framework for Cross-layer Attacks in Cognitive Radio Networks. In International Conference on Intelligent and Smart Computing in Data Analytics (pp. 23-33). Springer, Singapore. [22] Ganesh, Davanam, Thummala Pavan Kumar, and Malchi Sunil Kumar. "Optimised Levenshtein centroid cross‐layer defence for multi‐hop cognitive radio networks." IET Communications 15.2 (2021): 245-256. [23] Natarajan, V. Anantha, et al. "Segmentation of nuclei in histopathology images using fully convolutional deep neural architecture." 2020 International Conference on computing and information technology (ICCIT-1441). IEEE, 2020. [24] Sreedhar, B., BE, M. S., & Kumar, M. S. (2020, October). A comparative study of melanoma skin cancer detection in traditional and current image processing techniques. In 2020 Fourth International Conference on I-SMAC (IoT in Social, Mobile, Analytics and Cloud)(I-SMAC) (pp. 654-658). IEEE. Communications on Applied Nonlinear Analysis ISSN: 1074-133X Vol 31 No. 6s (2024) 439 https://internationalpubls.com [25] Ganesh, D., Kumar, T. P., & Kumar, M. S. (2021). Optimised Levenshtein centroid cross‐layer defence for multi‐hop cognitive radio networks. IET Communications, 15(2), 245-256. [26] Balaji, K., P. Sai Kiran, and M. Sunil Kumar. "Resource aware virtual machine placement in IaaS cloud using bio-inspired firefly algorithm." Journal of Green Engineering 10 (2020): 9315- 9327. [27] Peneti, S., Sunil Kumar, M., Kallam, S., Patan, R., Bhaskar, V., & Ramachandran, M. (2021). BDN-GWMNN: internet of things (IoT) enabled secure smart city applications. Wireless Personal Communications, 119(3), 2469-2485. [28] Balaji, K., P. Sai Kiran, and M. Sunil Kumar. "Power aware virtual machine placement in IaaS cloud using discrete firefly algorithm." Applied Nanoscience (2022): 1-9. [29] Davanam, G., Kumar, T. P., & Kumar, M. S. (2021). Efficient energy management for reducing cross layer attacks in cognitive radio networks. Journal of Green Engineering, 11, 1412-1426. [30] Kumar, M. Sunil, and K. Jyothi Prakash. "Internet of things: IETF protocols, algorithms and applications." Int. J. Innov. Technol. Explor. Eng 8.11 (2019): 2853-2857. [31] AnanthaNatarajan, V., Kumar, M. S., & Tamizhazhagan, V. (2020). Forecasting of Wind Power using LSTM Recurrent Neural Network. Journal of Green Engineering, 10. [32] Rupesh, B., & Kumar, M. S. (2015). Predicting the Hard Keyword Queries over Relational Databases. International Journal of Applied Engineering Research, 10(10), 26629-26640.