Inspection and Verification of 3D Laser Scanning Datasets Registration for Reverse Engineering Application, Part 1: A Product with Datum
M Muslimin (Tokyo Institute of Technology, United Kingdom (Great Britain))
Registration of point cloud datasets from a laser scanner is crucial in pursuing an accurate product in the reverse engineering. In practice, dataset error and registration error are the main sources of failure in achieving an accurate registration result. The lack of the inspection method of registration result becomes a challenge in guaranteeing the accuracy. This paper presents a novel method of the inspection and verification of point clouds registration to minimize the registration errors. In this method, aligned point clouds are inspected and verified using data from a CMM taken in correspondence with pre-defined object features. Point cloud segmentation and feature extraction are utilized to define concerned features while a grid method is used to perform the correspondence between the considered feature and CMM dataset. The method was implemented into a real case product to analyze its reliability. Based on the implementation, the proposed method showed the effectiveness in inspecting and verifying the registration result.
Journal: International Journal of Simulation- Systems, Science and Technology- IJSSST V20
Published: Feb 28, 2019