DETECTION OF DELAMINATION AND PEELING ON INFRASTRUCTURE SURFACES BY TIME SERIES ANALYSIS AND 3D FEATURE EXTRACTION OF LASER 3D POINT CLOUD DATA

Abstract

 Aging infrastructure stocks are important social issues. In this article, a mobile mapping system was utilized to realize automatic and realtime detection of delamination and peeling on tunnel concrete surfaces and potholes on road surfaces from laser 3D point cloud data. Time series analysis was applied to estimate profiles of tunnels and roads. After signal and image processing, candidate areas of damages were localized. Tunnel appendages were eliminated considering 3D features to visualize damages in 3D maps. Validity of the algorithm was demonstrated by real tunnels and roads, achieving about 80% detection rate of delamination and peeling on tunnel linings.