Research on Pavement Disease Detection Method Based on Improved Mask R-CNN

Aiming at the issues of insufficient pavement disease feature information extraction, low detection accuracy and inaccurate anchor box positioning in the detection process in road surface disease detection, this paper proposes a pavement disease detection method fused with Swin Transformer based on the Mask R-CNN detection model. This method uses Swin Transformer to replace its feature extraction network to obtain more effective feature information. And the K-means++ clustering algorithm is used to redesign the size of the anchor box to solve the problem of inaccurate anchor positioning in the detection process. The experimental results show that the detection accuracy of the proposed method is 5.81 0/0 higher than that of the original method, which effectively improves the ability to detect road surface diseases.