Advertisement

A new texture analysis approach

  • Zhang Zhan
  • Hou Jiechang
  • Liao Mengyang
Article
  • 15 Downloads

Abstract

A new texture analysis approach by using Gibbs random field model is discussed. A parameter vector is defined in order to compute potentials of textures. It is found that texture features can be provided by its corresponding parameter vector, and the vector can be estimated in terms of maximum potential. Conclusions can be inferred that this new method based on the maximum potential is valid.

Key words

texture Gibbs random field potential parameter vector 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Cross G R, Jain A K. Markov random field texture model.IEEE Trans on PAMI, 1983, PAMI-5:25–39Google Scholar
  2. 2.
    Cohen F S, Cooper D B. Simple parallel hierarchical and relaxation algorithms for segmenting noncausal markovian random field.IEEE Trans on PAMI, 1987, PAMI-9: 195–219Google Scholar
  3. 3.
    Acuna C O. Texture modeling using Gibbs distributions.Graphic Models & Image Processing, 1992,54(3):210–222CrossRefGoogle Scholar
  4. 4.
    Levitan E, Chan M, Herman GT. Image modeling Gibbs priors.Graphic Models & Image Processing, 1995,57 (2):117–130CrossRefGoogle Scholar

Copyright information

© Springer 1998

Authors and Affiliations

  • Zhang Zhan
    • 1
  • Hou Jiechang
    • 1
  • Liao Mengyang
    • 1
  1. 1.College of Electronic InformationWuhan UniversityWuhanChina

Personalised recommendations