<![CDATA[Rice is a major food commodity that is susceptible to leaf diseases, such as blast, bacterial leaf blight, and tungro, which can significantly reduce productivity if not detected early. This study aims to develop an early detection method for rice leaf diseases using a deep learning approach with a VGG16-based Convolutional Neural Network (CNN) architecture. The data used came from the Rice Leaf Dataset (Kaggle) and field images in Pringsewu Regency. The training process was carried out through transfer learning. The results showed that the model was able to achieve an accuracy of 99.75% on the training data, 96.08% on the validation data, and 100% on the test data. Field tests also proved the model's ability to generalize to real conditions, although there were still some cases with prediction confidence levels that were close between classes. These findings confirm that VGG16-based CNNs are effective for accurate and efficient detection of rice leaf diseases. The application of this model has the potential to support faster decision-making, reduce pesticide use, and encourage environmentally friendly sustainable agricultural practices.]]>
Copyrights © 2025
