Abstract: Traditional underwater inspection methods like manual diving inspection and underwater video recording are no longer applicable to the underwater inspection in the Yangtze River due to large flow rate and muddy water. In this research we adopted multi-beam sounding technique to perform full-coverage scanning to obtain high-precision point cloud data. Through 3D modeling, we visualized the conditions of underwater buildings and quantitatively analyzed the defects of water intake pipes of a water plant in the Yangtze River as a case study. Using the Kongsberg EM2040C (dual-probe) multi-beam sounding system, we designed the measurement route and chose reasonable wave speed and launch angle, and precisely processed the obtained high-precision and high-resolution underwater point cloud data using adaptive clip. The result accurately reflects the status of broken water pipes and broken locations, and provides reliable basic data for the water plant’s intake restoration project. The achievements can also be taken as reference for the condition check and reinforcement and restoration of similar projects.

Key words: multi-beam sounding technology, underwater inspection, adaptive filter, point cloud data, quantitative analysis of defects, 3D modeling