Motion Segmentation by Multibody Trifocal Tensor Using Line Correspondence

We present a geometric approach to multibody motion segmentation from line correspondences. Given three perspective views containing multiple moving objects, we demonstrate that after applying a polynomial embedding to the line correspondences, they became related by the so-called multibody trilinear constrain and its associated multibody trifocal tensor. We show how to obtain the individual trifocal tensor from the second order derivatives of the multibody trilinear constraint. Given trifocal tensors, we obtain the clustering of the motions and correspondences. Experimental results on synthetic and real dynamic scenes are presented

[1]  Bernhard P. Wrobel,et al.  Multiple View Geometry in Computer Vision , 2001 .

[2]  René Vidal,et al.  Optical flow estimation & segmentation of multiple moving dynamic textures , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[3]  René Vidal,et al.  A Unified Algebraic Approach to 2-D and 3-D Motion Segmentation , 2004, ECCV.

[4]  Jianhua Wang,et al.  Motion segmentation of multiple translating objects using line correspondences , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[5]  René Vidal,et al.  The multibody trifocal tensor: motion segmentation from 3 perspective views , 2004, Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2004. CVPR 2004..

[6]  René Vidal,et al.  A closed form solution to direct motion segmentation , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).