<![CDATA[Rice cultivation plays a crucial role in national food security, but is often hampered by leaf disease attacks that significantly impact production decline. To address this challenge, this study designed an application based on the Convolutional Neural Network (CNN) algorithm to classify rice leaf diseases automatically and accurately. Data collection was conducted through direct observation at Gapoktan (Farmer Group Association) in Kuningan Regency, interviews with farmers, literature studies, and system development using the Rapid Application Development (RAD) approach, which enables rapid and structured application development. The CNN model was tested using 480 sample images and achieved a high accuracy of 97.75%. The F1-Score values obtained were 0.97 for Brown Spot, 0.921 for Blast, 0.871 for Hispa, and 0.952 for healthy leaves. These results indicate that the application is capable of early disease detection, enabling farmers to take immediate preventive measures to minimize crop losses. To improve performance, it is recommended to apply optimization techniques to the CNN model, such as dataset expansion, various dataset augmentation techniques, and evaluation of high-complexity images. Development into other disease classifications is also highly potential. Overall, this application has significant potential to support digital agriculture and realize a more sustainable and modern rice farming system.]]>
Copyrights © 2026
