Classification of Plant Diseases Based on Rice Leaf Image Analysis Using SVM and CLAHE Methods

Abstract

Rice disease has been found in Indonesia and needs to be identified early for the prevention process. Blast and brown spot disease in rice is considered the most prominent and dangerous disease. This study will focus on identifying four rice leaf disease detections, including Bacterial Blight, Blast, Brown Spot and Tungro. This study aims to determine the effect of contrast enhancement method on SVM method for plant disease classification based on rice leaf image analysis using SVM and CLAHE methods. The dataset consists of four classes namely: Bacterial Blight, Blast, Brown Spot and Tungro with .jpg format. The dataset consists of 480 data for each class. Based on the experimental results, the accuracy of the SVM model reached 100% for the training stage and 94.81% for the testing stage. The accuracy of the CLAHE-SVM model reaches 100% for the training stage and 95.95% for the testing stage. Based on the accuracy value, the CLAHE-SVM model has better performance than the SVM model