Crop Disease Detection Based on YOLOv11

Traditional crop disease detection relies on manual labor, which is inefficient and limited in accuracy. In order to improve the efficiency and accuracy of crop detection, the model Yolo11-FBS is designed based on the Yolo11 model, which is improved by faterblock to C3k2, lightweight while improving performance. And the neck network is replaced with the network structure based on BIFPN modification, which strengthens the before and after feature map connection and enhances its feature extraction ability. Finally, efficient downsampling is realized by SCDown to reduce its computational and parametric quantities, and the results show that Yolo11-FBS achieves mAP50 and mAP50-95 scores of 0.788 and 0.672, which are improved by $2.3 \%$ and $2.4 \%$, respectively, compared with the YOLO11 model. It also decreased the number of parameters by $\mathbf{1 7. 4 \%}$ and GFLOPs by 0.3.