2 January 2018 The registration of non-cooperative moving targets laser point cloud in different view point
Author Affiliations +
Abstract
Non-cooperative moving target multi-view cloud registration is the key technology of 3D reconstruction of laser threedimension imaging. The main problem is that the density changes greatly and noise exists under different acquisition conditions of point cloud. In this paper, firstly, the feature descriptor is used to find the most similar point cloud, and then based on the registration algorithm of region segmentation, the geometric structure of the point is extracted by the geometric similarity between point and point, The point cloud is divided into regions based on spectral clustering, feature descriptors are created for each region, searching to find the most similar regions in the most similar point of view cloud, and then aligning the pair of point clouds by aligning their minimum bounding boxes. Repeat the above steps again until registration of all point clouds is completed. Experiments show that this method is insensitive to the density of point clouds and performs well on the noise of laser three-dimension imaging.
© (2018) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Shuai Wang, Shuai Wang, Huayan Sun, Huayan Sun, Huichao Guo, Huichao Guo, } "The registration of non-cooperative moving targets laser point cloud in different view point", Proc. SPIE 10456, Nanophotonics Australasia 2017, 104565G (2 January 2018); doi: 10.1117/12.2282651; https://doi.org/10.1117/12.2282651
PROCEEDINGS
6 PAGES


SHARE
RELATED CONTENT


Back to Top