The Deep Learning Based Analysis and Prediction of Apple Black Rot Disease in Apple Plants using ResNet-50 Hybrid Algorithm

Apple black rot, caused by the fungus Botryosphaeria obtusa, significantly impacts apple trees, causing notable crop damage. This disease appears as black, sunken spots on fruits, leaves, and branches. Early signs include small, purple spots on leaves that expand and darken. Infected fruits develop concentric decay rings, known as "frogeye" spots. The fungus survives winter in mummified fruits and dead wood, spreading via wind and rain. Effective control involves removing and destroying infected material, using appropriate fungicides, and maintaining orchard cleanliness. Prompt detection and intervention are essential to prevent severe outbreaks and maintain healthy apple yields. This study introduces a deep-learning classifier designed to distinguish between two types of image data, one showing apple leaves affected by apple black rot disease and the other showing healthy apple leaves. The images undergo de-noising to improve quality. The cleaned data is then used to train ResNet-50 networks, achieving high precision. In summary, this research aims to develop a highly accurate image classifier for detecting apple black rot disease, achieving an accuracy rate of 95.86%.