<![CDATA[Diseases affecting rice plants are one of the major factors contributing to decreased agricultural productivity and potential losses for farmers. Conventional disease identification generally relies on expert knowledge and is often impractical to perform accurately and efficiently in the field. This study aims to develop an image-based classification system for rice leaf diseases using a Deep Learning approach with a Convolutional Neural Network architecture, specifically MobileNetV2. The dataset consists of five rice leaf condition classes, namely bacterial disease, brown spot, blast, tungro, and healthy leaves, obtained from the Roboflow platform. The research methodology includes data collection, image pre-processing, model training using a transfer learning approach, and performance evaluation. Experimental results demonstrate that the proposed MobileNetV2 model achieves an accuracy of 93.46% and shows strong performance across most disease categories. Although misclassification still occurs among classes with similar visual characteristics, the results indicate that the developed model has significant potential as an efficient and automated decision-support system for rice plant disease identification.]]>
Copyrights © 2026
