Automatic Pipeline Generation by the Sequential Segmentation and Skelton Construction of Point Cloud

Abstract

Point clouds are the main output of automatic data collection using laser-scanner technologies. The large numbers of points scanned from pipeline plants make the plant reconstruct very difficult. We have developed a system based on the Point cloud Library (PCL), to find, recognize and reconstruct 3D pipes within point-clouds fully automatically. The proposed approach consists of pre-processing point cloud data, segmentation, skeleton extraction and automatic cylinder fitting for pipeline. The presented results shows that the proposed method enables reliable 3D models of pipelines, which could be successfully incorporated into the reconstruction of a plant information modelling method and utilized for, assist maintenance and expansion of existing plants.