Apple Leaf Disease Detection Algorithm Based on an Improved YOLOv8

To improve the precision of recognizing apple leaf diseases in real-world applications, a detection model for apple leaf diseases is proposed, utilizing an enhanced YOLOv8 architecture. A Convolutional Block Attention Module (CBAM), incorporating both channel and spatial attention mechanisms, is introduced to enhance the extraction of disease-specific features while effectively ignoring irrelevant information in the images. A DyHead detection head replaces the original head in the YOLOv8 model, integrating scale, spatial, and task attention mechanisms to further enhance the model's detection performance. Compared with the original YOLOv8 algorithm, the improved YOLOv8 model increases [email protected] by 1.6% and [email protected]:0.95 by 1.2%. Experimental results demonstrate that the proposed improved YOLOv8 algorithm achieves high recognition accuracy and detection speed, providing a precise real-time detection method for apple leaf diseases.