YOLO11-DL: A Lightweight Eggplant Disease Object Detection Network

With the development of smart agriculture, accurate recognition of eggplant leaf diseases is crucial for improving crop yield and quality. However, various factors such as different angles, lighting, and disease area size in real planting environments affect the performance of object detection. To address this, we propose an eggplant leaf disease recognition algorithm named YOLO11DL, aimed at achieving precise identification and analysis. First, a large-scale dataset containing 11,693 images was constructed, covering 6 common diseases, and data augmentation techniques were applied to enhance the diversity and robustness of the dataset. Next, the YOLOv11n backbone network was improved with the introduction of the C3K2_DWR module to enhance multi-scale information capture efficiency and high-level feature extraction ability, ensuring accurate disease recognition and classification. Additionally, a lightweight detection head, LSCSBD, was designed, utilizing shared convolution layers and independent BN layers to reduce model parameters while maintaining high detection accuracy. Experimental results show that the improved YOLOv11n_DL model achieved a [email protected] of 0.837 on the eggplant leaf disease dataset, with a 6.9% reduction in parameters and a computational cost of only 6.4 GFLOPs, outperforming mainstream models like YOLOv8 and YOLOv9t. The algorithm is suitable for deployment on mobile devices, providing farmers with a fast and accurate disease recognition tool to support intelligent agricultural management. Future work will further expand the dataset and explore additional optimization strategies.