Key points are not available for this paper at this time.
In the early diseases of asphalt pavement, cracks detection is very important, which can provide support for the follow-up highway maintenance work. This paper adopts the YOLOv5 algorithm of target detection in deep learning, and takes the road image data sets of seven cities as an example to detect pavement cracks. Firstly, the data set is transformed and split, then trained and predicted, and finally the results are analyzed. Through comparison, the algorithm has achieved good detection results.
Wu et al. (Fri,) studied this question.