Concerning Bayesian motion segmentation, model averaging, matching and the trifocal tensor

  • P. H. S. Torr
  • A. Zisserman
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1406)


Motion segmentation involves identifying regions of the image that correspond to independently moving objects. The number of independently moving objects, and type of motion model for each of the objects is unknown a priori.

In order to perform motion segmentation, the problems of model selection, robust estimation and clustering must all be addressed simultaneously. Here we place the three problems into a common Bayesian framework; investigating the use of model averaging-representing a motion by a combination of models—as a principled way for motion segmentation of images. The final result is a fully automatic algorithm for clustering that works in the presence of noise and outliers.


Mixture Model Fundamental Matrix Motion Segmentation Posterior Model Probability Trifocal Tensor 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


  1. 1.
    H. Akaike. A new look at the statistical model identification. IEEE Trans. on Automatic Control, Vol. AC-19(6):716–723, 1974.MathSciNetCrossRefGoogle Scholar
  2. 2.
    M. Armstrong, A. Zisserman, and R. Hartley. Self-calibration from image triplets. In Proc. 4th ECCV, pages 3–16. Springer-Verlag, 1996.Google Scholar
  3. 3.
    S. Ayer and H. Sawhney. Layered representation of motion video using robust maximum-likelihood estimation of mixture models and mdl encoding. In Proc. 5th ICCV, pages 777–784, 1995.Google Scholar
  4. 4.
    P. Beardsley, P. H. S. Torr, and A. Zisserman. 3D model aquisition from extended image sequences. In 4th ECCV, pages 683–695. Springer-Verlag, 1996.Google Scholar
  5. 5.
    A. Black and P. Anandan. The robust estimation of multiple motions: Parametric and piecewise-smooth flow fields. CVIU, 63(1):75–104, 1996.Google Scholar
  6. 6.
    M. Brooks, L. Agapito, D. Huynh, and L. Baumela. Direct methods for self-calibration of a moving head. In Proc. 4th ECCV, pages 97–108. Springer-Verlag, 1996.Google Scholar
  7. 7.
    T. Darrell, S. Sclaroff, and A. Pentland. Segmentation by minimal description. In Proc. 3rd ICCV, pages 112–116. IEEE, 1990.Google Scholar
  8. 8.
    A. P. Dempster, N. M. Laird, and D. B. Rubin. Maximum likelihood from incomplete data via the em algorithm. J. R. Statist. Soc., 39 B:1–38, 1977.MathSciNetGoogle Scholar
  9. 9.
    D. Draper. Assessment and propagation of model uncertainty (with discussion). Journal of the Royal Statistical Society series B, 57:45–97, 1995.zbMATHMathSciNetGoogle Scholar
  10. 10.
    O.D. Faugeras. What can be seen in three dimensions with an uncalibrated stereo rig? In G. Sandini, editor, Proc. 2nd ECCV, pages 563–578. Springer-Verlag, 1992.Google Scholar
  11. 11.
    R. A. Fisher Uncertain Inference Proc. Amer. Acad. Arts and Sciences, 71 245–258, 1936.zbMATHGoogle Scholar
  12. 12.
    M. Fischler and R. Bolles. Random sample consensus: a paradigm for model fitting with application to image analysis and automated cartography. Commun. Assoc. Comp. Mach., vol. 24:381–95, 1981.MathSciNetGoogle Scholar
  13. 13.
    E. FranÇois and P. Bouthemy. Multiframe based identification of mobile components of a scene with a moving camera. In Proc. CVPR., 1991.Google Scholar
  14. 14.
    E. Francois and P. Bouthemy. Derivation of qualitative information in motion analysis. IVC, vol. 8(4):279–287, Nov 1990.CrossRefGoogle Scholar
  15. 15.
    H. Gu, S. Yoshiaki, and Asada M. MDL-based segementation and motion modeling in a long image sequence of scene with multiple independently moving objects. PAMI, 18:58–64, 1996.Google Scholar
  16. 16.
    C. Harris and M. Stephens. A combined corner and edge detector. In Proc. Alvey Conf., pages 189–192, 1987.Google Scholar
  17. 17.
    R. I. Hartley. Estimation of relative camera positions for uncalibrated cameras. In Proc. 2nd ECCV, pages 579–87. Springer-Verlag, 1992.Google Scholar
  18. 18.
    R. I. Hartley. Lines and points in three views — a unified approach. In ARPA Image Understanding Workshop, Monterey, 1994.Google Scholar
  19. 19.
    R. I. Hartley and P. Sturm. Triangulation. In American Image Understanding Workshop, pages 957–966, 1994.Google Scholar
  20. 20.
    J. S. Hodges. Uncertainty, policy analysis and statistics (with discussion). Statistical Science, 2:259–291, 1987.Google Scholar
  21. 21.
    H. Jeffreys. Theory of Probability. Clarendon Press, Oxford, third edition, 1961.Google Scholar
  22. 22.
    A. Jepson and M. Black. Mixture models for optical flow computation. In cvpr-93, pages 760–766. IEEE, 1993.Google Scholar
  23. 23.
    K. Kanatani. Automatic singularity test for motion analysis by an information criterion. In Proc. 4th ECCV, pages 697–708, Springer-Verlag, 1996.Google Scholar
  24. 24.
    K. Kanatani. Statistical Optimization for Geometric Computation: Theory and Practice. Elsevier Science, Amsterdam, 1996.Google Scholar
  25. 25.
    R. E. Kass and A. E. Raftery. Bayes factors. Journal of the American Statistical Association, 90:733–795, 1995.Google Scholar
  26. 26.
    M. Kendall and A. Stuart. The Advanced Theory of Statistics. Charles Griffin and Company, London, 1983.Google Scholar
  27. 27.
    E. E. Leamer. Specification searches: ad hoc inference with nonexperimental data. Wiley, New York, 1978.Google Scholar
  28. 28.
    D. Madigan and A. E. Raftery. Model selection and accounting for model uncertainty in graphical models using Occam's window. Journal of the American Statistical Association, 89:1535–1546, 1994.zbMATHCrossRefGoogle Scholar
  29. 29.
    G.I. McLachlan and K. Basford. Mixture models: inference and applications to clustering. Marcel Dekker. New York, 1988.Google Scholar
  30. 30.
    P. F. McLauchlan, I. Reid, and D. W. Murray. Coarse image motion for saccade control. In Proc. BMVC. Springer-Verlag, Sept 1992. Leeds.Google Scholar
  31. 31.
    P. F. McLauchlan, I. Reid, and D. W. Murray. Recursive affine structure and motion from image sequences. In Proc. 3rd ECCV, volume 1, pages 217–224, May 1994.Google Scholar
  32. 32.
    A. J. Miller. Selection of subsets of regression variables (with discussion). Journal of the Royal Statistical Society (Series A), 147:389–425, 1984.zbMATHGoogle Scholar
  33. 33.
    B. R. Moulton. A Bayesian-approach to regression selection and estimation with application to a price-index for radio services. Journal of Econometrics, 49:169–193, 1991.CrossRefGoogle Scholar
  34. 34.
    J. Mundy and A. Zisserman. Geometric Invariance in Computer Vision. MIT press, 1992.Google Scholar
  35. 35.
    R.C. Nelson. Qualitative detection of motion by a moving observer. IJCV, pages 33–46, 1991.Google Scholar
  36. 36.
    R. A. Redner and H. F. Walker. Mixture densities, maximum likelihood and the EM algorithm. SIAM Review, 26:195–239, 1984.zbMATHMathSciNetCrossRefGoogle Scholar
  37. 37.
    B. D. Ripley. Pattern recognition and neural networks. Cambridge University Press, Cambridge, 1996.Google Scholar
  38. 38.
    B. Russell. History of Western Philosophy. Routledge, 1961.Google Scholar
  39. 39.
    G. Schwarz. Estimating dimension of a model. Ann. Stat., 6:461–464, 1978.zbMATHGoogle Scholar
  40. 40.
    G.A.F. Wild C. J. Seber. Non-Linear Regression. Wiley, New York, 1989.Google Scholar
  41. 41.
    L. S. Shapiro. Affine Analysis of Image Sequences. PhD thesis, Oxford University, 1993.Google Scholar
  42. 42.
    A. Shashua. Trilinearity in visual recognition by alignment. In Proc. 3rd ECCV, volume 1, pages 479–484, May 1994.MathSciNetGoogle Scholar
  43. 43.
    M. Spetsakis and J. Aloimonos. A multi-frame approach to visual motion perception. IJCV, 6:245–255, 1991.CrossRefGoogle Scholar
  44. 44.
    C. V. Stewart. Minpran; a new robust estimator for computer vision. IEEE Trans. PAMI, vol.PAMI-17, no.10:925–938, 1995.Google Scholar
  45. 45.
    P. H. S. Torr. Outlier Detection and Motion Segmentation. PhD thesis, Dept. of Engineering Science, University of Oxford, 1995.Google Scholar
  46. 46.
    P. H. S. Torr and D. W. Murray. The development and comparison of robust methods for estimating the fundamental matrix. Int Journal of Computer Vision, 24(3):271–300, 1997.CrossRefGoogle Scholar
  47. 47.
    P. H. S. Torr and A Zisserman. Robust parameterization and computation of the trifocal tensor. Image and Vision Computing, 15:591–607, 1997.CrossRefGoogle Scholar
  48. 48.
    P. H. S. Torr, A Zisserman, and S. Maybank. Robust detection of degenerate configurations for the fundamental matrix. To Appear CVIU, 1998.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1998

Authors and Affiliations

  • P. H. S. Torr
    • 1
    • 2
  • A. Zisserman
    • 1
    • 2
  1. 1.Microsoft ResearchRedmondUSA
  2. 2.Dept. of Engineering ScienceOxford UniversityEngland

Personalised recommendations