Three-Dimensional Point Cloud Registration Based on Normal Vector Angle
- PDF / 1,566,052 Bytes
- 9 Pages / 595.276 x 790.866 pts Page_size
- 98 Downloads / 251 Views
RESEARCH ARTICLE
Three-Dimensional Point Cloud Registration Based on Normal Vector Angle Liang Li1 • Xingyan Cao1 • Jie Sun1 Received: 7 April 2018 / Accepted: 20 November 2018 Indian Society of Remote Sensing 2018
Abstract The widespread use of three-dimensional laser scanning makes point cloud registration essential for model reconstruction. Although iterative closest point (ICP) is a widely used algorithm for automatic and accurate point cloud registration, the ICP algorithm is time-intensive and typically falls in local optima. We employ the normal vector angle of the 3D model and integrate a heuristic search into the ICP algorithm to improve its efficiency, especially while registering complex surfaces. We compare this improved ICP algorithm with the original ICP algorithm and variants of the algorithm in terms of convergence, robustness, and response from a poor initial cloud point position to verify that the proposed improvements outperform the conventional ICP algorithm in the three evaluated aspects when processing point clouds with complex surfaces. Keywords 3D laser scanning Point cloud Fine registration Iterative closest point (ICP)
Introduction As lasers cannot penetrate an entity during 3D scanning, the information of the target surface must be obtained by measurements from multiple acquisition positions. Although the coordinate systems among scanning points differ, the measurements must be registered to the same coordinate system before processing and model reconstruction. Hence, the registration accuracy directly affects the quality of the reconstructed model, thus being an essential process for three-dimensional (3D) laser scanning. One of the most widely used algorithms for fine registration is iterative closest point (ICP), introduced by Besl and Mckay (1992). Basically, ICP creates corresponding points considering the minimum Euclidean distance between two datasets to calculate the rigid transformation matrix minimizing the root mean square (RMS) error between the datasets.
& Xingyan Cao [email protected] 1
School of Environment Science and Spatial Informatics, China University of Mining and Technology, Xuzhou 221116, Jiangsu, People’s Republic of China
Despite being automatic and highly accurate, the ICP algorithm is time- and memory-intensive. In addition, it has two major limitations: One dataset is a subset of the other, and the initial positions of the datasets should be close to prevent falling into a local optimum. Moreover, the ICP algorithm can easily match incorrect points, thus reducing the convergence speed. To address these shortcomings, algorithm improvements have been addressed regarding three key aspects: improving the constraints to speed up convergence; developing a search strategy to improve the efficiency while determining corresponding points; and improving the sampling method to reduce matching error. To minimize the error function, the ICP algorithm calculates the rigid transformation matrix using iterative least squares. To increase the convergence speed, Mi
Data Loading...
