ST-IEEFO: a hybrid optimization-based swin transformer for plant disease detection

In agriculture, identifying and diagnosing plant diseases have become a critical problem. Early plant disease detection helps implement preventive measures and overcome the spread of infections. Depending on the plant diseases can range from mild damage to complete destruction and making early detection essential to mitigate such losses. To address this issue, this work presents a multi-plant leaf disease detection using a Deep Learning (DL) approach Improved Electric Eel Foraging Optimization (IEEFO) assisted Swin Transformer (ST) and the model is named ST-IEEFO. The suggested model integrated the powerful feature extracting ability of the ST with optimization potential of the IEEFO. The study utilizes the Plant Village data and obtained better accuracy (99.8%), and recalls values (99.6%). Thus, the suggested model shows the betterment of the ST-IEEFO over other DL models. By effectively recognizing and classifying plant diseases, this approach ensures a better solution for early disease detection. Further, it contributes to sustainable agricultural practices.