<![CDATA[Chili peppers are a strategic horticultural commodity in Indonesia, but their productivity is often hampered by plant diseases that are difficult to identify accurately and in a timely manner using conventional methods. This study aims to develop an accurate and efficient system for classifying chili pepper plant diseases using Convolutional Neural Networks (CNN). The MobileNetV2 architecture was used due to its computational efficiency, making it suitable for implementation on devices with limited resources. A dataset containing 4,000 images of chili leaves was used, categorized into four classes: healthy leaves, leaves infected with fruit flies, curled leaves, and anthracnose. The model was trained and verified, achieving an overall accuracy of 89.62%. Evaluation results showed strong performance with precision of 89.93%, recall of 89.62%, and an F1 score of 89.54%. The trained model was successfully integrated into a web-based application to facilitate use by farmers and agricultural officers. The findings conclude that the proposed MobileNetV2-based system provides an effective and practical solution for early detection of chili diseases, supporting precision agriculture and regional food security initiatives in Morowali Regency.]]>
Copyrights © 2026
