Detection and Classification of Plant Disease Using Hybrid AI Deep Learning Techniques

To reduce financial losses and boost agricultural yield, plant leaf diseases must be accurately and promptly detected. However, farmers' dependence on traditional manual methods sometimes makes it difficult to diagnose certain illnesses with accuracy. The application of the the VGG16 model, also known as CNN, for the detection then identification of plant leaf diseases is investigated in this study. In order to create sophisticated disease prediction algorithms for agriculture, the study makes use of the extensive Plant Village A database, which contains more than 70,000 images of both healthy then sick leaves of crops from fourteen distinct species. Data augmentation methods including horizontal flipping and histogram equalization remained castoff to improve the dataset then increase the model's flexibility. The performance of the VGG16 model had been thoroughly assessed by contrasting it with well-known CNN designs including ResNet, Inception-V3 and VGG-19. With the help of the Plant Village dataset, proposed work produced an impressive 99.12% accuracy rate. With values near 0.99, evaluation criteria such as recall, precision, F1-score, then accuracy continuously showed excellent performance, confirming the effectiveness of the suggested strategy. The study's discoveries determine the usefulness of VGG16 as a potent tool for precise disease diagnosis and point to important developments in plant disease detection. Because they make it possible to use better crop protection and disease control techniques, these advancements are extremely valuable to all parties involved in agriculture, including farmers including researchers.