Hybrid-based segmentation of massive three-dimensional point cloud data

Data segmentation is an important part of the reverse engineering. Point cloud data segmentation is a technique of dividing different regions of the stitched scattered point cloud into a single geometric property. The technique is a critical part of the former reverse modeling. This article draws ona segmentation method based on hybrid, and proposes a grouping method of improving three-dimensional point cloud data which based on backtracking. During the measurement process, the method avoids the sparse phenomenon of point cloud cause by object occlusion, the change of reflecting rate or unsatisfactory reduction algorithm processing and some other circumstances. Thereby, it resolves the "isolated island"question of point cloud which generated in the split process.