Abstract: Water leakage always occurs in shield tunnels during operations that may result in cracks and corrosion of rebars. Such defects could jeopardize the safety of tunnel operations. Water leakage detection is thus necessary to ensure safe tunnel operations. This study proposes a method using 3D laser scanning technology for identifying locations of water leakages in shield tunnels. Firstly,the proposed method collects point cloud data for generating a grayscale image of the tunnel surface based on the corrected reflection intensity values. Then,dilation and erosion algorithms are used for preprocessing of the imagery data. Next,the authors use connected domain algorithm to derive locations and area of water leakage. Finally,the authors have validated the proposed method through a case study. Results show that the proposed method could achieve 92% accuracy for water leakage detection in shield tunnels.

Key words: water leakage detection, shield tunnel, 3D laser scanning, reflection intensity correction, grayscale image