Research on intelligent detection of urban road subgrade disease based on deep learning algorithm

Urban road collapse has the characteristics of concealment and sudden subgrade defects in the early stage, so it is necessary to identify and investigate subgrade collapse diseases in the early stage. In order to comprehensively study subgrade disease the research results of domestic and foreign scholars mainly focus on the intelligent identification of subgrade disease caused by urban road collapse. By deeply understanding the basic concept of deep learning and the basic principle of target detection algorithm, an algorithm suitable for detecting subgrade disease caused by city road collapse is selected. In this paper, the image database of roadbed collapse is established, and some images in the database are used to train various target detection algorithms, and the training results of various algorithms are analyzed. According to the test index, the optimal algorithm is selected. Finally, the optimized algorithm is tested by using the processed actual geological radar image, and the feasibility of its application in road detection is observed. The automatic identification of the geological radar image is realized by the computer program, and the detection accuracy is better.