Stabilizing Image Mosaicing by Model Selection

  • Yasushi Kanazawa
  • Kenichi Kanatani
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2018)


The computation for image mosaicing using homographies is numerically unstable and causes large image distortions if the matching points are small in number and concentrated in a small region in each image. This instability stems from the fact that actual transformations of images are usually in a small subgroup of the group of homographies. It is shown that such undesirable distortions can be removed by model selection using the geometric AIC without introducing any empirical thresholds. It is shown that the accuracy of image mosaicing can be improved beyond the theoretical bound imposed on statistical optimization. This is made possible by our knowledge about probable subgroups of the group of homographies.We demonstrate the effectiveness of our method by real image examples.


Model Selection Real Image Rigid Motion Image Transformation Outdoor Scene 
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.
    H. Akaike,A newlook at the statistical model identification, IEEE Trans. Automation Control, 19-6 (1974), 716–723.CrossRefMathSciNetGoogle Scholar
  2. 2.
    W. Förstner, Reliability analysis of parameter estimation in linear models with applications to mensuration problems in computer vision, Comput. Vision Graphics Image Process., 40 (1987), 273–310.CrossRefGoogle Scholar
  3. 3.
    K. Kanatani, Geometric Computation for Machine Vision, Oxford University Press, Oxford, 1993.zbMATHGoogle Scholar
  4. 4.
    K. Kanatani, Statistical Optimization for Geometric Computation: Theory and Practice, Elsevier Science, Amsterdam, 1996.zbMATHCrossRefGoogle Scholar
  5. 5.
    K. Kanatani, Geometric information criterion for model selection, Int. J. Comput. Vision, 26-3 (1998), 171–189.CrossRefGoogle Scholar
  6. 6.
    K. Kanatani, Statistical optimization and geometric inference in computer vision, Phil. Trans. Roy. Soc. Lond., A-356 (1998), 1303–1320.CrossRefMathSciNetGoogle Scholar
  7. 7.
    K. Kanatani and N. Ohta, Accuracy bounds and optimal computation of homography for image mosaicing applications, Proc. 7th Int. Conf. Comput. Vision, September, 1999, Kerkya, Greece, pp. 73–78.Google Scholar
  8. 8.
    C. Matsunaga and K. Kanatani, Calibration of a moving camera using a planar pattern: Optimal computation, reliability evaluation and stabilization by model selection, Proc. 6th Euro. Conf. Comput. Vision, June-July 2000, Dublin, Ireland, Vol.2, pp. 595–609.Google Scholar
  9. 9.
    D. D. Morris and T. Kanade, A unified factorization algorithm for points, line segments and planes with uncertainty models, Proc. Int. Conf. Comput. Vision, January 1998, Bombay, India, pp. 696–702.Google Scholar
  10. 10.
    H. S. Sawhney, S. Hsu and R. Kumar, Robust video mosaicing through topology inference and local to global alignment, Proc. 5th Euro. Conf. Comput. Vision, June 1998, Freiburg, Germany, Vol. 2, pp. 103–119.Google Scholar
  11. 11.
    J. Shi and C. Tomasi, Good features to track, Proc. Conf. Comput. Vision Patt. Recogn., June 1994, Seattle,WA, pp. 593–600.Google Scholar
  12. 12.
    A. Singh,An estimation-theoretic framework for image-flowcomputation, Proc. 3rd Int. Conf. Comput. Vision, December, dy1990, Osaka, Japan, pp. 168–177.Google Scholar
  13. 13.
    R. Szeliski and H.-U. Shum, Creating full view panoramic image mosaics and environment maps, Proc. SIGGRAPH’97, August 1997, Los Angeles, CA, U.S.A., pp. 251–258.Google Scholar
  14. 14.
    P. H. S. Torr, Model selection for two view geometry:A review, in, D. A. Forsyth, J. L. Mundy, V. D. Gesú, R. Cipolla (Eds.): Shape, Contour and Grouping in ComputerVision, LNCS 1681, Springer-Verlag, Berlin, 1999, pp. 277–301.CrossRefGoogle Scholar
  15. 15.
    T. Werner, T. Pajdla and V. Hlaváč, Efficient 3-D scene visualization by image extrapolation, Proc. 5th Euro. Conf. Comput. Vision, June 1998, Freiburg, Germany, Vol. 2, pp. 382–396.Google Scholar
  16. 16.
    I. Zoghlami, O. Faugeras and R. Deriche, Using geometric corners to build a 2D mosaic from a set of images, Proc. Conf. Comput. Vision Patt. Recogn., June 1997, Puerto Rico, pp. 420–425.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2001

Authors and Affiliations

  • Yasushi Kanazawa
    • 1
  • Kenichi Kanatani
    • 2
  1. 1.Department of Knowledge–based Information EngineeringToyohashi University of TechnologyAichiJapan
  2. 2.Department of Computer ScienceGunma UniversityGunmaJapan

Personalised recommendations