ISBN: 978-981-18-5852-9 DOI: 10.18178/wcse.2022.04.062
An Acceleration Method to Iterative Closet Point Registration Algorithm
Abstract— To solve the problem of low registration efficiency of the classical ICP algorithm in point cloud data registration, a point cloud registration method to improve the ICP algorithm is proposed. By considering the similarity of the point cloud data, our algorithm find a more reasonable point pair to compare with the classical ICP. In addition, we used a new distance error measure in the ICP framework, which considers the similarity instead of the Euclidean distance. By observing the experimental results, we find that the proposed scheme speeded up the registration convergence and eliminated the mispairing of data points. The running time of the algorithm is significantly improved to meet the requirements of the registration, and the Registration efficiency is improved to 8 times the original.
Index Terms— point cloud, registration, iterative closet point, distance error function, optimization, overlap, similarity
Miao Yang
Xi'an University of Posts and Telecommunications School of Computing
Yagang Wang
Xi'an University of Posts and Telecommunications School of Computing
Jungang Han
Xi'an University of Posts and Telecommunications School of Computing
Kaixuan Wang
Xi'an University of Posts and Telecommunications School of Computing
Cite: Miao Yang, YaGang Wang, JunGang Han, Kaixuan Wang, " An Acceleration Method to Iterative Closet Point Registration Algorithm, " WCSE 2022 Spring Event: 2022 9th International Conference on Industrial Engineering and Applications, pp. 527-533, Sanya, China, April 15-18, 2022.