Automatic view planning for 3D reconstruction and occlusion handling based on the integration of active and passive vision

In order to acquire three-dimensional model of the unknown object automatically, we proposed a novel view planning method, which combines the active and passive vision and determines viewpoints similar to human intuition. Firstly, it gives the concept of limit visual surface in binocular vision system, and the next best view is obtained with the predicted contour. Secondly, for the occlusion in reconstruction process, the occluded contour can be predicted by SFS (Shape from Shading) or limit visual surface. Finally, an added viewpoint is selected with the goal to fill the occlusion in range image, which makes the reconstructed model complete. Experimental result is presented to illustrate the effectiveness and versatility of our approach.

[1]  Marc Pollefeys,et al.  3D Occlusion Inference from Silhouette Cues , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.

[2]  Fillia Makedon,et al.  Automatic sensor placement in a 3D volume , 2009, PETRA '09.

[3]  Zhengyou Zhang,et al.  A Flexible New Technique for Camera Calibration , 2000, IEEE Trans. Pattern Anal. Mach. Intell..

[4]  C. Ian Connolly,et al.  The determination of next best views , 1985, Proceedings. 1985 IEEE International Conference on Robotics and Automation.

[5]  Xiaobu Yuan,et al.  A Mechanism of Automatic 3D Object Modeling , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  Hui Ding,et al.  A Novel Occlusion Planning Method for Unknown 3D Objects Automatic Reconstruction , 2010, 2010 International Conference on Artificial Intelligence and Computational Intelligence.

[7]  Christophe Dumont,et al.  A next-best-view system for autonomous 3-D object reconstruction , 2000, IEEE Trans. Syst. Man Cybern. Part A.

[8]  David W. Capson,et al.  Subspace position measurement in the presence of occlusion , 2002, Pattern Recognit. Lett..

[9]  Mubarak Shah,et al.  A fast linear shape from shading , 1992, Proceedings 1992 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.

[10]  Xiangyang Ju,et al.  Self-correction of 3D reconstruction from multi-view stereo images , 2009, 2009 IEEE 12th International Conference on Computer Vision Workshops, ICCV Workshops.

[11]  Richard Pito,et al.  A Solution to the Next Best View Problem for Automated Surface Acquisition , 1999, IEEE Trans. Pattern Anal. Mach. Intell..

[12]  Berthold K. P. Horn.  SHAPE FROM SHADING: A METHOD FOR OBTAINING THE SHAPE OF A SMOOTH OPAQUE OBJECT FROM ONE VIEW , 1970 .

[13]  Youfu Li,et al.  Research on new view planning method for automatic reconstruction of unknown 3D objects , 2009, 2009 IEEE International Conference on Robotics and Biomimetics (ROBIO).