Research on Methods of Pavement Distress Detection Based on Object Detection Convolutional Neural Network

Aiming at the problem that the accuracy of pavement distress detection is easily affected by complex texture, noisy background, uneven illumination conditions and external environmental interference of highway pavement distress images, this paper studies the methods of pavement distress detection based on object detection Convolutional Neural Network. Firstly, the methods of pavement distress detection based on Faster-RCNN, YOLO-V5s, and SSD are compared and analyzed. Secondly, three variants of CNN models are investigated for pavement distress detection, including FR-PDD, YOLO5s-PDD and SSD-PDD. Finally, the comparative experiments were conducted, and the results showed that the average of Yolov5s-PDD network is superior to the other two methods, with an average accuracy of 98.1%.