Detection of Diseases in Rubber Leaves Based on Different Classification Models and Comparatives Analysis Using Python

This paper explores the process of identifying foliar diseases by analyzing the environment in which they grow. The study was conducted on 4000 rubber leaf samples and peripheral extraction and contour segmentation techniques were used to determine the classification level of diseased leaves and their infection rates. Support Vector Machine, Random Forest, Decision Tree, and Confusion Matrix models have been used for leaf disease detection. Based on this absolute classification, it automatically acquires the ability to extract very useful features. The SVM technique produced the most accurate method for the dataset with a success rate of 97.87%. The results of the conducted experiments exhibit the superiority of SVM algorithm for rubber leaf disease detection as compared to all other classification techniques. Visual-based methods for detecting foliar diseases also provide satisfactory results and demonstrate excellent operational efficiency.