Multi-source point cloud registration method based on automatically calculating overlap

被引：0
|
作者：
Li X. [1 ]
Mo S. [1 ]
Huang H. [1 ]
Yang S. [1 ]
机构：
[1] College of Electrical Engineering, Sichuan University, Chengdu
来源：
Hongwai yu Jiguang Gongcheng/Infrared and Laser Engineering | 2021年 / 50卷 / 12期
关键词：
Contribution factor; Improved TrICP; Point cloud registration; Singular value decomposition;
D O I：
10.3788/IRLA20210088
中图分类号：
学科分类号：
摘要：
In order to solve the problems (noise, partial overlap, registration parameters determining for different models, etc.) in multi-source point cloud registration, an improved TrICP algorithm based on the contribution factor was proposed. First, the improved voxel down-sampling and random down-sampling methods were adopted to resample the point cloud. The contribution factor was proposed to investigate the point pairs that contributed more to the registration. The transformation matrix was solved by using singular value decomposition. At the same time, slopes between points on the distance curve and the original point were used to calculate the overlap automatically. Therefore, the automatic registration of point cloud was realized. Comparative experiments among several registration algorithms were conducted based on the Stanford University Bunny point cloud and the 'Maoxian 624' landslide point cloud. The results show that the speeds of the improved algorithm on Bunny and landslide increase by 50 % and 67 % respectively and the accuracies are improved. In addition, it performs well even with a lot of noise. The improved algorithm can align the laser point cloud and point cloud from image reconstruction effectively and automatically, which contain lots of noise, partial overlap, non-homology. Then, the advantages of multi-source data are combined to obtain the accurate point cloud information of the target. Copyright ©2021 Infrared and Laser Engineering. All rights reserved.
引用
收藏
相关论文
共 20 条
  • [1] Du Ruijian, Ge Baozhen, Chen Lei, Texture mapping of multi-view high-resolution images and binocular 3D point clouds, Chinese Optics, 13, 5, pp. 1055-1064, (2020)
  • [2] Qiao Yujing, Jia Baoming, Jiang Jingang, Et al., Networking method of multi-view stereo-vision measurement network, Infrared and Laser Engineering, 49, 7, (2020)
  • [3] Besl P J, Mckay N D., A method for registration of 3-D shapes, IEEE Transactions on Pattern Analysis and Machine Intelligence, 14, 2, pp. 239-256, (1992)
  • [4] Wang Bin, Liu Lin, Hou Yuqing, Et al., Three-dimensional cardiac point registration by improved iterative closet point method, Optics and Precision Engineering, 28, 2, pp. 474-484, (2020)
  • [5] Wang Shuai, Sun Huayan, Guo Huichao, Overlapping region extraction method for laser point clouds registration, Infrared and Laser Engineering, 46, (2017)
  • [6] Chetverikov D, Stepanov D, Krsek P., Robust euclidean alignment of 3D point sets: the trimmed iterative closest point algorithm, Image and Vision Computing, 23, 3, pp. 299-309, (2005)
  • [7] Dai Jinglan, Chen Zhiyang, Ye Xiuzi, The application of ICP algorithm in point ploud alignment, Journal of Image and Graphics, 12, 3, pp. 517-521, (2007)
  • [8] Zhong Ying, Zhang Meng, Automatic registration technology of point cloud based on improved ICP algorithm, Control Engineering of China, 21, 1, pp. 37-40, (2014)
  • [9] Zuo Chao, Lu Min, Tan Zhiguo, Et al., A novel algorithm for registration of point clouds, Chinese Journal of Lasers, 39, 12, (2012)
  • [10] Chen Jie, Cai Yong, Zhang Jiansheng, Point cloud registration based on entropy criterion genetic algorithm, Application Research of Computers, 36, 1, pp. 316-320, (2019)
12