Advertisement

Non-parametric local transforms for computing visual correspondence

  • Ramin Zabih
  • John Woodfill
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 801)

Abstract

We propose a new approach to the correspondence problem that makes use of non-parametric local transforms as the basis for correlation. Non-parametric local transforms rely on the relative ordering of local intensity values, and not on the intensity values themselves. Correlation using such transforms can tolerate a significant number of outliers. This can result in improved performance near object boundaries when compared with conventional methods such as normalized correlation. We introduce two non-parametric local transforms: the rank transform, which measures local intensity, and the census transform, which summarizes local image structure. We describe some properties of these transforms, and demonstrate their utility on both synthetic and real data.

Keywords

Normalize Correlation Local Intensity Correspondence Problem Stereo Algorithm Scene Element 
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.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    P. Aschwanden and W. Guggenbühl. Experimental results from a comparative study on correlation-type registration algorithms. In Förstner and Ruwiedel, editors, Robust Computer Vision, pages 268–289. Wichmann, 1993.Google Scholar
  2. 2.
    Paul Besl, Jeffrey Birch, and Layne Watson. Robust window operators. In International Conference on Computer Vision, pages 591–600, 1988.Google Scholar
  3. 3.
    Michael Black and P Anandan. A framework for the robust estimation of optical flow. In International Conference on Computer Vision, pages 231–236, 1993.Google Scholar
  4. 4.
    U. Dhond and J. Aggarwal. Structure from stereo — a review. IEEE Transactions on Systems, Man and Cybernetics, 19(6), 1989.Google Scholar
  5. 5.
    Stuart Geman and Donald Geman. Stochastic relaxation, gibbs distributions, and the bayesian restoration of images. IEEE PAMI, 6:721–741, 1984.Google Scholar
  6. 6.
    Marsha Jo Hanna. Computer Matching of Areas in Stereo Images. PhD thesis, Stanford, 1974.Google Scholar
  7. 7.
    R. Hodgson, D. Bailey, M. Naylor, A. Ng, and S. McNeill. Properties, implementations and applications of rank filters. Journal of Image and Vision Computing, 3(1):3–14, February 1985.Google Scholar
  8. 8.
    Michael Kass. Computing visual correspondence. DARPA Image Understanding Proceedings, pages 54–60, 1983.Google Scholar
  9. 9.
    E. L. Lehman. Nonparametrics: statistical methods based on ranks. Holden-Day, 1975.Google Scholar
  10. 10.
    David Marr and Keith Nishihara. Representation and recognition of the spatial organization of three-dimensional shapes. Proceedings of the Royal Society of London B, 200:269–294, 1978.Google Scholar
  11. 11.
    Peter Meer, Doron Mintz, Azriel Rosenfeld, and Dong Yoon Kim. Robust regression methods for computer vision: A review. International Journal of Computer Vision, 6(1):59–70, 1991.Google Scholar
  12. 12.
    H. Keith Nishihara. Practical real-time imaging stereo matcher. Optical Engineering, 23(5):536–545, Sept–Oct 1984.Google Scholar
  13. 13.
    Tomaso Poggio, Vincent Torre, and Christof Koch. Computational vision and regularization theory. Nature, 317:314–319, 1985.Google Scholar
  14. 14.
    Peter Seitz. Using local orientational information as image primitive for robust object recognition. SPIE proceedings, 1199:1630–1639, 1989.Google Scholar
  15. 15.
    Ramin Zabih. Individuating Unknown Objects by Combining Motion and Stereo. PhD thesis, Stanford University, 1994 (forthcoming).Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1994

Authors and Affiliations

  • Ramin Zabih
    • 1
  • John Woodfill
    • 2
  1. 1.Computer Science DepartmentCornell UniversityIthacaUSA
  2. 2.Interval Research CorporationPalo AltoUSA

Personalised recommendations