Wide Baseline Point Matching Using Affine Invariants Computed from Intensity Profiles

  • Dennis Tell
  • Stefan Carlsson
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1842)


The problem of establishing correspondences between images taken from different viewpoints is fundamental in computer vision. We propose an algorithm which is capable of handling larger changes in viewpoint than classical correlation based techniques. Optimal performance for the algorithm is achieved for textured objects which are locally planar in at least one direction. The algorithm works by computing affinely invariant fourier features from intensity profiles in each image. The intensity profiles are extracted from the image data between randomly selected pairs of image interest points. Using a voting scheme, pairs of interest points are matched across images by comparing vectors of fourier features. Outliers among the matches are rejected in two stages, a fast stage using novel view consistency constraints, and a second, slower stage using RANSAC and fundamental matrix computation. In order to demonstrate the quality of the results, the algorithm is tested on several different image pairs.


Feature Vector Interest Point Query Image Visual Servoing Point Correspondence 
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.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    R. Basri, E. Rivlin, I. Shimshoni, Visual homing: Surfing on the epipoles. ICCV 1998, pp 863–869.Google Scholar
  2. 2.
    R. Basri, E. Rivlin, I. Shimshoni, Image-based robot navigation under the perspective model. ICRA 1999, pp 2578–2583.Google Scholar
  3. 3.
    T. Tuytelaars, L. vanGool et. al, Matching of affinely invariant regions for visual servoing. ICRA 1999, pp 1601–1606.Google Scholar
  4. 4.
    D. Tell, S. Carlsson, View based visual servoing using epipolar geometry. SCIA 1999, pp 63–70.Google Scholar
  5. 5.
    Z. Zhang, R. Deriche, O. Faugeras, Q.-T. Luong, A Robust technique for matching two uncalibrated images through the recovery of the unknown epipolar geometry. AI Journal, vol 78, pp 87–119, 1995.Google Scholar
  6. 6.
    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, pp 381–395, 1981.Google Scholar
  7. 7.
    P. Pritchett, A. Zisserman, Wide baseline stereo matching. ICCV 1998, pp 863–869.Google Scholar
  8. 8.
    C. Schmid and R. Mohr, Local greyvalue invariants for image retrieval. PAMI, vol 19, no 5, pp 872–877, 1997.Google Scholar
  9. 9.
    C. Schmid, R. Mohr and C. Bauckhage Comparing and evaluating interest points. ICCV 1998.Google Scholar
  10. 10.
    A. Jain, Fundamentals of digital image processing, Prentice Hall, 1989.Google Scholar
  11. 11.
    S. Carlsson, Recognizing walking people, ECCV 2000.Google Scholar
  12. 12.
    Z. Zhang, Software for computing the fundamental matrix. Web page
  13. 13.
    Z. Zhang, Determining the epipolar geometry and its uncertainty: a review. IJCV 1997, vol 27, no 2, pp 161 195, 1998.CrossRefGoogle Scholar
  14. 14.
    M. de Berg, M. vanKreveld, et. al, Computational Geometry: Algorithms and Applications. Springer-Verlag, 1997.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2000

Authors and Affiliations

  • Dennis Tell
    • 1
  • Stefan Carlsson
    • 1
  1. 1.Royal Institute of Technology (KTH)CVAP/NADAStockholmSweden

Personalised recommendations