A Point Cloud Registration Algorithm Based on 3D-SIFT

Point cloud registration is a key technology in reverse engineering. The registration process of point cloud is divided into coarse registration and fine registration. For fine registration process, ICP (Iterative Close Point) is a classic algorithm. The traditional ICP algorithm is inefficient and incorrect if the correct initial point set is not obtained. In this paper, a point cloud registration algorithm based on 3D-SIFT features is proposed. In this method, the 3D-SIFT algorithm is used to extract key points. At the same time, the 3D-feature descriptor is used as a constraint on the initial set of points in the ICP algorithm. The results show that the method improves the efficiency and precision of the ICP algorithm, and achieves better results of point cloud registration.

[1]  Gérard G. Medioni,et al.  Object modelling by registration of multiple range images , 1992, Image Vis. Comput..

[2]  Paul J. Besl,et al.  A Method for Registration of 3-D Shapes , 1992, IEEE Trans. Pattern Anal. Mach. Intell..

[3]  Kostas Daniilidis,et al.  Fully Automatic Registration of 3D Point Clouds , 2006, 2006 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'06).

[4]  Anath Fischer,et al.  3D Point Cloud Registration for Localization Using a Deep Neural Network Auto-Encoder , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[5]  Chaomin Luo,et al.  Generic object recognition based on the fusion of 2D and 3D SIFT descriptors , 2015, 2015 18th International Conference on Information Fusion (Fusion).

[6]  Gary K. L. Tam,et al.  Registration of 3D Point Clouds and Meshes: A Survey from Rigid to Nonrigid , 2013, IEEE Transactions on Visualization and Computer Graphics.

[7]  Jie Zhao,et al.  Distributed randomized singular value decomposition using count sketch , 2017, 2017 International Conference on Security, Pattern Analysis, and Cybernetics (SPAC).

[8]  Nico Blodow,et al.  Fast Point Feature Histograms (FPFH) for 3D registration , 2009, 2009 IEEE International Conference on Robotics and Automation.

[9]  Wang Jun,et al.  LiDAR point cloud registration based on improved ICP method and SIFT feature , 2015, 2015 IEEE International Conference on Progress in Informatics and Computing (PIC).

[10]  Wen-Chung Chang,et al.  Eye-in-hand vision-based robotic bin-picking with active laser projection , 2015, The International Journal of Advanced Manufacturing Technology.