<![CDATA[This study developed a web-based application for the classification and detection of tomato leaf diseases using Convolutional Neural Network (CNN) and You Only Look Once (YOLO) models. The research followed a Research and Development approach that consisted of requirement analysis, system design, implementation, model training, and testing. The CNN model was trained to classify tomato leaf images into specific disease categories, while the YOLO model was designed to detect and localize diseased areas in real time. Both models were integrated into a Flask-based web system to provide accessible and interactive functionality through standard web browsers. Testing results showed that the CNN model achieved an accuracy of 96.1%, effectively identifying disease types such as Early Blight and Bacterial Spot. The YOLO model reached a mean Average Precision (mAP) of 87.3% for real-time detection, successfully locating and labeling infected regions on tomato leaves. The integration of CNN and YOLO models demonstrated strong classification and detection performance, offering an efficient and scalable solution to support early disease diagnosis and digital transformation in precision agriculture.]]>
Copyrights © 2026
