Automating Rice Crop Disease Classification with VGG16 and Deep Learning

Rice is a basic crop for almost half of the world's population, various diseases like Bacterial Blight, Blast, Brown Spot, and False Smut impair its output. Early, correct detection of these illnesses guarantees both effective management and lowest possible crop losses. The study proposes a fine-tuned VGG16 deep learning model strategy for rice crop diseases classification. It consists of 200 high quality images encompassing the four disease categories. Adding custom layers to focus on the specific job of disease classification and freezing the first layers helped the model to transformed for the fine-tuned vgg16 model. The fine-tuned VGG16 model attained an accuracy of 94%. Performance metrics including precision, recall, and F1-score showed outstanding results especially for Brown Spot and False Smut infections. The study provides a strong tool for early disease detection, thereby enabling sustainable management of rice fields and maybe higher agricultural yield.