Recovering image structure by model-based interaction map

  • Georgy Gimel'farb
Document Image Analysis and Recognition
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1451)


We address modelling of stochastic image textures by Gibbs random fields with a translation invariant structure of multiple pairwise pixel interactions. The characteristic interaction structure and strengths (Gibbs potentials) are learnt from a given training sample by analytic and stochastic approximation of the unconditional or conditional maximum likelihood estimates of the potentials. The interaction structure is revealed by a model-based interaction map showing the relative contributions of each interaction to a total Gibbs energy. Features of the interaction maps are discussed and illustrated by experiments with various natural textures.


Interaction Structure Image Patch Stochastic Approximation Texture Type Pixel Pair 
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.
    Besag, J. E.: Spatial interaction and the statistical analysis of lattice systems. J. Royal Statistical Soc. B36 (1974) 192–236Google Scholar
  2. 2.
    Brodatz, P.: Textures: A Photographic Album for Artists an Designers. New York: Dover Publications (1966)Google Scholar
  3. 3.
    Chellappa, R., Jain, A. (Eds): Markov Random Fields: Theory and Application. Academic Press (1993)Google Scholar
  4. 4.
    Chetverikov, D.: On some basic concepts of texture analysis. Proc. 2nd Int. Conf. on Computer Analysis of Images and Patterns. Sept. 2–4, 1987, Wismar, GDR. Berlin: Akademie-Verlag (1987) 196–201Google Scholar
  5. 5.
    Chetverikov, D., Haralick, R. M.: Texture anisotropy, symmetry, regularity: recovering structure and orientation from interaction maps. Proc. 6th British Machine Vision Conf. September 11–14, 1995, Birmingham. Sheffield: Univ. of Sheffield (1995) 57–66.Google Scholar
  6. 6.
    Gimel'farb, G. L.: Texture modeling by multiple pairwise pixel interactions. IEEE Trans. Pattern Anal. Machine Intell. 18:11 (1996) 1110–1114CrossRefGoogle Scholar
  7. 7.
    Gimel'farb, G. L.: Non-Markov Gibbs texture model with multiple pairwise pixel interactions. Proc. 13th IAPR Int. Conf. Pattern Recognition, vol. II, Aug.25–29, 1996, Vienna, Austria. TUWien (1996) 591–595Google Scholar
  8. 8.
    Gimel'farb, G. L.: Analytic approximation of Gibbs potentials to model stochastic textures. Proc. First Joint Australia & New Zealand Biennial Conf. on Digital Image & Vision Computing: Techniques and Applications (DICTA'97 & IVCNZ'97)Albany, Auckland, New Zealand, 10–12 Dec. 1997. Massey Univ.: Palmerston North, 1997, 153–158.Google Scholar
  9. 9.
    Gimel'farb G. L., Jain, A. K.: On retrieving textured images from an image data base. Pattern Recognition 29:9 (1996) 1461–1483CrossRefGoogle Scholar
  10. 10.
    Haralick, R. M., Shapiro, L. G.: Computer and Robot Vision. Vol. 1. Reading:Addison-Wesley Publ. Co. (1992) 453–507Google Scholar
  11. 11.
    Julesz, B.: Textons, the elements of texture perception, and their interactions. Nature 290 (1981) 91–97CrossRefPubMedGoogle Scholar
  12. 12.
    Zalesny, A. V.: Homogeneity & texture. General approach. Proc. 12th IAPR Int. Conf. on Pattern Recognition. October 9–13, 1994, Jerusalem, Israel. Vol. 1. Los Alamitos: IEEE Comp. Soc. Press (1994) 592–594.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1998

Authors and Affiliations

  • Georgy Gimel'farb
    • 1
  1. 1.CITR, Department of Computer Science, Tamaki CampusThe University of AucklandAucklandNew Zealand

Personalised recommendations