Design and Development of Machine Learning Approaches for Tomato Leaf Disease Identification and Categorization

Agriculture holds immense significance in India. Various diseases manifest in several varieties of crops and vegetables. The tomato holds significant importance as a vegetable in India. Tomatoes are vulnerable to several diseases that decrease crop production, such as bacterial spot, late blight, septoria leaf spot, and yellow bent leaf. This study presents a module that use automated classification techniques to identify disease symptoms in plant leaves. Currently, plant diseases pose a significant challenge in the field of industrial agriculture. Untreated plant diseases can result in substantial economic losses if not detected promptly. The objective of this study is to discern and categorize diseases in tomato plants. The process of detecting ailments in tomatoes is arduous, exceedingly subjective, and challenging. Photographs of tomato diseases suffer from a deficiency of visual differentiation. Consequently, it might provide a challenge to discern a tomato leaf just based on an image without performing feature extraction and using convolution for the purpose of classification. This study examines and evaluates the efficacy of several techniques for extracting features and classifying them to identify illnesses in tomato leaves. The approaches encompass GLCM, Gabor, SURF, SVM, KNN, and Naive Bayes. As a result of the model we developed, feature extraction and classification can now be executed entirely automatically. Through the utilisation of various performance metrics to assess the system's performance, we successfully identified the optimal feature set and classification methodology for the purpose of disease classification in tomato leaves. The experimental findings indicate that the Gabor characteristics have the capability to precisely detect and classify various diseases affecting tomato leaves. SVM is more accurate than alternative classification algorithms, despite its delayed execution time. Early disease detection can potentially reduce farmers' dependence on pesticides.