Automated extraction of tunnel leakage location and area from 3D laser scanning point clouds

Abstract

Water leakage is one of the common tunnel disease problems. Fast and accurate detection is crucial for timely management of leakage disease. This paper introduces a method for automatically extracting the location and area of tunnel leakage from 3D laser scanning point clouds. The method involves three key steps: segmenting the point cloud data along the tunnel axis based on the width of the tube sheet, correcting the intensity of the point cloud using a linear intensity correction model, and binarizing the point cloud data to extract the leakage area and calculate its area. To evaluate the performance of the proposed methodology, water leakage detection software is prepared based on the method, and tests are carried out in the urban subway tunnels in the leakage disease-producing sections. The results show a detection accuracy of 92 % and an improvement in detection efficiency of 7–8 times. The method proposed can be used to quickly and accurately extract the location and area of water leakage from point cloud data.