Automated 3D scan multi-view registration based on rotation estimation

{huaxin.wang, j.s.m.vergeest, y.song, t.wiegers}@tudelft.nl ABSTRACT A new approach is presented for automatic registration of a sequence of 3D scanned range images into a complete model. Provided that the user follows the proposed scanning strategy, the difficulty to register the split views into a complete model is alleviated and success rate of automatic registration is improved by an adaptive forward estimation based on the extracted rotation axes and angles of the object between each scanning steps. The basic idea of this improvement is that when each of the scan differs slightly from its previous one in a small rotation about an approximately constant axis, the rotation axes and angle in the previous scanning step can be used as the first estimation of the rotation in the next step, so as to avoid manual coarse initial alignment. Formulation of the principle is provided and experimental results demonstrated.

[1]  Anthony J. Maeder,et al.  The correspondence framework for 3D surface matching algorithms , 2005, Comput. Vis. Image Underst..

[2]  Yonghuai Liu,et al.  Improving ICP with easy implementation for free-form surface matching , 2004, Pattern Recognit..

[3]  Sang Wook Lee,et al.  Invariant features and the registration of rigid bodies , 1999, Proceedings 1999 IEEE International Conference on Robotics and Automation (Cat. No.99CH36288C).

[4]  Joaquim Salvi,et al.  Range image registration for industrial inspection , 2005, IS&T/SPIE Electronic Imaging.

[5]  Andrew E. Johnson,et al.  Using Spin Images for Efficient Object Recognition in Cluttered 3D Scenes , 1999, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  Heinz Hügli,et al.  Multi-feature matching algorithm for free-form 3D surface registration , 1998, Proceedings. Fourteenth International Conference on Pattern Recognition (Cat. No.98EX170).

[7]  Naokazu Yokoya,et al.  A Robust Method for Registration and Segmentation of Multiple Range Images , 1995, Comput. Vis. Image Underst..

[8]  Yu Song,et al.  ICP fitness analysis for 3D scan data matching , 2006 .

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

[10]  John J. Craig Zhu,et al.  Introduction to robotics mechanics and control , 1991 .

[11]  Yonghuai Liu,et al.  Automatic 3d free form shape matching using the graduated assignment algorithm , 2005, Pattern Recognit..

[12]  Reinhard Klein,et al.  Image-Based Registration of 3D-Range Data Using Feature Surface Elements , 2004, VAST.

[13]  Martial Hebert,et al.  Fully automatic registration of multiple 3D data sets , 2003, Image Vis. Comput..

[14]  Paolo Cignoni,et al.  Exploiting the scanning sequence for automatic registration of large sets of range maps , 2005, Comput. Graph. Forum.