Registration for 3-D point cloud using angular-invariant feature

This paper proposes an angular-invariant feature for 3-D registration procedure to perform reliable selection of point correspondence. The feature is a k-dimensional vector, and each element within the vector is an angle between the normal vector and one of its k nearest neighbors. The angular feature is invariant to scale and rotation transformation, and is applicable for the surface with small curvature. The feature improves the convergence and error without any assumptions about the initial transformation. Besides, no strict sampling strategy is required. Experiments illustrate that the proposed angular-based algorithm is more effective than iterative closest point (ICP) and the curvature-based algorithm.

[1]  Yoel Shkolnisky,et al.  The angular difference function and its application to image registration , 2005, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[2]  Ruigang Yang,et al.  Registering, integrating, and building CAD models from range data , 1998, Proceedings. 1998 IEEE International Conference on Robotics and Automation (Cat. No.98CH36146).

[3]  Mel Siegel,et al.  Remote Inspection Technologies for Aircraft Skin Inspection , 1997 .

[4]  Stefano Soatto,et al.  Integral Invariant Signatures , 2004, ECCV.

[5]  Jun Jiang,et al.  A registration method based on angular-invariant feature , 2008, 2008 2nd International Symposium on Systems and Control in Aerospace and Astronautics.

[6]  Sang Wook Lee,et al.  ICP Registration Using Invariant Features , 2002, IEEE Trans. Pattern Anal. Mach. Intell..

[7]  Marc Levoy,et al.  Efficient variants of the ICP algorithm , 2001, Proceedings Third International Conference on 3-D Digital Imaging and Modeling.

[8]  Ray Jarvis,et al.  3D free-form surface registration and object recognition , 2004, International Journal of Computer Vision.

[9]  Leonidas J. Guibas,et al.  Robust global registration , 2005, SGP '05.

[10]  Andrew E. Johnson,et al.  Surface landmark selection and matching in natural terrain , 2000, Proceedings IEEE Conference on Computer Vision and Pattern Recognition. CVPR 2000 (Cat. No.PR00662).

[11]  David Zhang,et al.  Three-dimensional surface registration: A neural network strategy , 2006, Neurocomputing.

[12]  Berthold K. P. Horn,et al.  Closed-form solution of absolute orientation using unit quaternions , 1987 .

[13]  David Fofi,et al.  A review of recent range image registration methods with accuracy evaluation , 2007, Image Vis. Comput..

[14]  Reinhard Koch,et al.  Invariant-based registration of surface patches , 1999, Proceedings of the Seventh IEEE International Conference on Computer Vision.

[15]  K. S. Arun,et al.  Least-Squares Fitting of Two 3-D Point Sets , 1987, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[16]  Igor Guskov,et al.  Multi-scale features for approximate alignment of point-based surfaces , 2005, SGP '05.

[17]  Gilbert Soucy.  Surface Recovery from Range Images Using Curvature and Motion Consistency , 1997, Comput. Vis. Image Underst..

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

[19]  Igor Guskov,et al.  Robust Alignment of Multi-view Range Data to CAD Model , 2006, IEEE International Conference on Shape Modeling and Applications 2006 (SMI'06).

[20]  Kari Pulli,et al.  Multiview registration for large data sets , 1999, Second International Conference on 3-D Digital Imaging and Modeling (Cat. No.PR00062).

[21]  Tamal K. Dey,et al.  Eurographics Symposium on Point-based Graphics (2005) Normal Estimation for Point Clouds: a Comparison Study for a Voronoi Based Method , 2022 .

[22]  Tony DeRose,et al.  Surface reconstruction from unorganized points , 1992, SIGGRAPH.