Motion Curves for Parametric Shape and Motion Estimation

  • Pierre-Louis Bazin
  • Jean-Marc Vézien
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2351)


This paper presents a novel approach to camera motion parametrization for the structure and motion problem. In a model-based framework, the hypothesis of (relatively) continuous and smooth sensor motion enables to reformulate the motion recovery problem as a global curve estimation problem on the camera path. Curves of incremental complexity are fitted using model selection to take into account incoming image data. No first estimate guess is needed. The use of modeling curves lead to a meaningful description of the camera trajectories, with a drastic reduction in the number of degrees of freedom. In order to characterize the behaviour and performances of the approach, experiments with various long video sequences, both synthetic and real, are undertaken. Several candidate curve models for motion estimation are presented and compared, and the results validate the work in terms of reconstruction accuracy, noise robustness and model compacity.


Structure from motion camera modeling model selection motion curves model-based estimation 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    P. L. Bazin. A parametric scene reduction algorithm from geometric relations. In Proc. Vision Geometry IX, SPIE’s 45th annual meeting, San Diego, 2000.Google Scholar
  2. 2.
    P. L. Bazin and J. M. Vézien. Tracking geometric primitives in video streams. In Proc. Irish Machine Vision and Image Processing Conference, Belfast, 2000.Google Scholar
  3. 3.
    Kishore Bubna and Charles V. Stewart. Model selection and surface merging in reconstruction algorithms. In Proc. International Conference on Computer Vision, Bombay, 1998.Google Scholar
  4. 4.
    Alessandro Chiuso, Paolo Favaro, Hailin Jin, and Stefano Soatto. 3-d motion and structure from 2-d motion causally integrated over time: Implementation. In Proc. European Conference on Computer Vision, Dublin, 2000.Google Scholar
  5. 5.
    A. Gelman, J. B. Carlin, H. S. Stern, and D. B. Rubin. Bayesian data analysis. Chapman & Hall, 1995.Google Scholar
  6. 6.
    Xiaoping Hu and Narendra Ahuja. Motion and structure estimation using long sequence motion models. Image and Vision Computing, 11(9):549–569, november 1993.Google Scholar
  7. 7.
    Frederik Kahl and Anders Heyden. Auto-calibration and euclidean reconstruction from continuous motion. In Proc. International Conference on Computer Vision, Vancouver, 2001.Google Scholar
  8. 8.
    K. Kanatani. Statistical Optimization for Geometric Computation: Theory and Practice. Elsevier Science, Amsterdam, 1996.zbMATHCrossRefGoogle Scholar
  9. 9.
    Kenichi Kanatani and Chikara Matsunaga. Geometric mdl and its media applications. In Proc. Workshop on Information-Based Induction Sciences, Izu, 2000.Google Scholar
  10. 10.
    S. J. Maybank and P. Sturm. Minimum description length and the inference of scene structure from images. In Proc. IEE Colloquium on Applications of Statistics to Pattern Recognition, 1999.Google Scholar
  11. 11.
    P. H. S. Torr. Data segmentation and model selection for computer vision, chapter Model selection for structure and motion recovery from multiple images. Springer, 2000.Google Scholar
  12. 12.
    P. Vitányi and M. Li. Ideal mdl and its relation to bayesianism. In Proc. Information, Statistics and Induction in Science Conference, Melbourne, 1996.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2002

Authors and Affiliations

  • Pierre-Louis Bazin
    • 1
  • Jean-Marc Vézien
    • 1
  1. 1.INRIA RocquencourtLe Chesnay CedexFrance

Personalised recommendations