Abstract

The point clouds scanned by a 3D laser scanner may be affine transformed when the size and posture of the objects being scanned are different. This type of problem is common, but few algorithms can solve it. Therefore, this Letter proposes a parallel registration algorithm. The algorithm eliminates the effects of the affine matrix in the point cloud, based on a simple whitening operation. Moreover, it also has strong anti-noise performance. The algorithm proposed in this Letter is not only simple in structure, but also shows excellent effects in practical applications and simulations.

© 2020 Chinese Laser Press

ICP registration with DCA descriptor for 3D point clouds

Ying He, Jun Yang, Xingming Hou, Shiyan Pang, and Jia Chen
Opt. Express 29(13) 20423-20439 (2021)

Multimodal affine registration for ICGA and MCSL fundus images of high myopia

Gaohui Luo, Xinjian Chen, Fei Shi, Yunzhen Peng, Dehui Xiang, Qiuying Chen, Xun Xu, Weifang Zhu, and Ying Fan
Biomed. Opt. Express 11(8) 4443-4457 (2020)

PDC-Net: Robust point cloud registration using deep cyclic neural network combined with PCA

Dengzhi Liu, Yu Zhang, Lin Luo, Jinlong Li, and Xiaorong Gao
Appl. Opt. 60(11) 2990-2997 (2021)

Improving the precision of parallel registration by incorporating a priori information

Nikolaus Berlakovich, Martin Fuerst, Ernst Csencsics, and Georg Schitter
Opt. Express 30(23) 41473-41491 (2022)

Iterative parallel registration of strongly misaligned wavefront segments

Nikolaus Berlakovich, Ernst Csencsics, Martin Fuerst, and Georg Schitter
Opt. Express 29(21) 33281-33296 (2021)