Research on apple disease recognition method based on deep learning

This study investigates the identification of apple leaf diseases using deep learning methodologies. Focusing on eight prevalent diseases—Alternaria blotch, brown spot, bacterial wilt, gray spot, mosaic virus disease, powdery mildew, rust, and scab—along with healthy leaves, a dataset of 13,775 images was compiled. A YOLOv8 deep learning framework was employed to train a model for apple disease recognition, enabling accurate classification of apple leaf diseases. Furthermore, a user-friendly, UI-based intelligent diagnostic system for apple leaf diseases was developed using the trained model and PyQT5. This system supports image, batch image, video, and camera-based detection, offering efficient and precise identification of various leaf diseases. The model achieves an average confidence level of 0.93 with an average detection time of 0.021 seconds, thereby significantly enhancing the quality and efficiency of agricultural production.