Multibody structure from motion could be solved by the factorization approach. However, the noise measurements would make the segmentation difficult when analyzing the shape interaction matrix. This paper presents an orthogonal subspace decomposition and grouping technique to approach such a problem. We decompose the object shape spaces into signal subspaces and noise subspaces. We show that the signal subspaces of the object shape spaces are orthogonal to each other. Instead of using the shape interaction matrix contaminated by noise, we introduce the shape signal subspace distance matrix for shape space grouping. Outliers could be easily identified by this approach. The robustness of the proposed approach lies in the fact that the shape space decomposition alleviates the influence of noise, and has been verified with extensive experiments.


    Access

    Check access

    Check availability in my library

    Order at Subito €


    Export, share and cite



    Title :

    Multibody grouping via orthogonal subspace decomposition


    Contributors:
    Ying Wu, (author) / Zhengyou Zhang, (author) / Huang, T.S. (author) / Lin, J.Y. (author)


    Publication date :

    2001-01-01


    Size :

    698160 byte





    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English



    Multibody Grouping via Orthogonal Subspace Decomposition

    Wu, Y. / Zhang, Z. / Huang, T. S. et al. | British Library Conference Proceedings | 2001


    Multibody Grouping from Motion Images

    Gear, C. W. | British Library Online Contents | 1998


    Photometric subspace for multibody motion segmentation

    Maki, A. | British Library Online Contents | 2004



    Illumination Subspace for Multibody Motion Segmentation

    Maki, A. / Hattori, H. / IEEE | British Library Conference Proceedings | 2001