Tunnel Water Leakage Detection And Crack Identification Based On 3D Laser Scanning Technology

As one of the most important components of rail transit and highways,ensuring the safety of tunnel operation is related to the national economy and people’s livelihood and the safety of people’s lives and property.However,superinticial diseases such as tunnel leakage and cracks affect the safety of tunnel operation,and the traditional timeconsuming and labor-intensive manual inspection methods can no longer meet the growing demand for tunnel inspection,and it is urgent and significant to study efficient and reliable tunnel nondestructive testing methods.At present,tunnel disease detection is only based on image data,which has problems such as complex tunnels,large environmental impact,complex detection process,and low detection accuracy.In view of the above problems,based on the intensity image data generated by laser point cloud and point cloud collected by 3D laser scanning technology,according to the respective advantages of laser point cloud and image data,this paper uses point cloud reflection intensity,deep learning and image processing technology to conduct leakage water detection and crack identification research on tunnels,and the main research contents and results are as follows:(1)The experimental results show that the proposed tunnel axis extraction method combining principal component analysis(PCA)and random sampling consistency(RANSAC)algorithm is proposed,and compared with the traditional method of least squares extraction of tunnel axis,the experimental results show that the proposed method not only reduces the error caused by human intervention,but also improves the accuracy and stability of tunnel axis extraction,lays a foundation for subsequent tunnel expansion along the central axis,and ensures the accuracy of tunnel leakage detection.(2)The tunnel water leakage detection method based on point cloud reflection intensity is studied,the segmentation threshold of the water leakage area is determined by using the histogram of the point cloud reflection intensity distribution,the method of extracting the water leakage area by Euclidean clustering and then using the Alpha Shapes algorithm for boundary extraction is designed,and finally the rapid detection and statistical analysis of tunnel water leakage disease is realized through MATLAB programming.(3)The tunnel crack recognition method combining Faster R-CNN and regional growth algorithm is studied,the improved Faster R-CNN model is used to detect and identify tunnel cracks,and the background gray compensation algorithm overcomes the influence of debris such as tunnel pipeline facilities on crack extraction,and then combines the regional growth algorithm to realize the identification and extraction of tunnel fine cracks,and the experimental results show that the method can quickly identify and successfully extract tunnel fine cracks and calculate their length and number.