Abstract: Using the pipe laser scanning technology, 3D information of point clouds on the inner surface of the pipeline can be obtained, which solves the problem that traditional pipeline video detection methods are difficult to quantify. Aiming at the characteristics of point cloud data of pipeline damage defects, point cloud filtering algorithms such as statistical filtering and radius filtering are used to achieve point cloud deletion at pipeline damage locations. Based on RANSAC algorithm, cylindrical surface projection of pipeline point cloud data and 2D planar expansion of pipeline point cloud data are implemented. Through grid filtering and region growth algorithms, pipeline point cloud data damage defect detection and automated analysis are achieved. The measured data verify the reliability and effectiveness of the method, which can accurately identify the location of pipeline damage, and calculate parameters such as damage area, longitudinal length, and circumferential length.

Key words: pipeline laser detection, point cloud data processing, automatic identification, damage detection, RANSAC