Research on detection of Pavement Surface Cracks Based on Improved YOLO v11

Pavement surface crack detection is crucial for road safety. To improve the accuracy and robustness of current crack detection methods, this paper proposes a YOLO v11-based model that incorporates dynamic snake module, CAFM attention mechanism, and SIoU loss function. Experimental results demonstrate improvements of $2.71 \%, 4.66 \%$, and $7.65 \%$ in [email protected], accuracy, and recall, respectively, compared to the original model. The proposed method ensures both real-time performance and accuracy, meeting the requirements for pavement crack detection.