Unsupervised image segmentation using Markov Random Field models

  • S. A. Barker
  • P. J. W. Rayner
Markov Random Fields
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1223)


We present an unsupervised segmentation algorithm based on a Markov Random Field model for noisy images. The algorithm finds the the most likely number of classes, their associated model parameters and generates a corresponding segmentation of the image into these classes. This is achieved according to the MAP criterion. To facilitate this, an MCMC algorithm is formulated to allow the direct sampling of all the above parameters from the posterior distribution of the image. To allow the number of classes to be sampled, a reversible jump is incorporated into the Markov Chain. The jump enables the possible splitting and combining of classes and consequently, their associated regions within the image. Experimental results are presented showing rapid convergence of the algorithm to accurate solutions.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [1]
    B.S.Manjunath and R.Chellappa. Unsupervised texture segmentation using Markov Random Fields. IEEE Trans. Patt. Anal & Machine Intell., 13(5):478–482, May 1991.Google Scholar
  2. [2]
    C.Kervrann and F.Heitz. A Markov Random Field model-based approach to unsupervised texture segmentation using local and global statistics. IEEE Trans. Image Processing, 4(6):856–862, June 1995.Google Scholar
  3. [3]
    C.S.Won and H.Derin. Unsupervised Segmentation of Noisy and Textured Images using Markov Random Fields. CVGIP:Graphical Models and Image Processing, 54(4):308–328, Jult 1992.Google Scholar
  4. [4]
    J.W.Modestino D.A.Langan and J.Zhang. Cluster Validation for Unsupervised Stochastic model-based Image Segmentation. Proc. ICIP94, pages 197–201, 1994.Google Scholar
  5. [5]
    C.Graffigne D.Geman, S.Geman and P.Dong. Boundary Detection by Constrained Optimization. IEEE Trans. Patt. Anal & Machine Intell., 12(7):609–628, July 1990.Google Scholar
  6. [6]
    D.K.Panjwani and G.Healey. Markov Random Field models for unsupervised segmentation of textured color images. IEEE Trans. Patt. Anal & Machine Intell., 17(10):939–954, Oct 1995.Google Scholar
  7. [7]
    F.S.Cohen and Z.Fan. Maximum Likelihood unsupervised texture segmentation. CVGIP:Graphical Models and Image Processing, 54(3):239–251, May 1992.Google Scholar
  8. [8]
    H.H.Nguyen and P.Cohen. Gibbs Random Fields, Fuzzy Clustering, and the unsupervised segmentation of images. CVGIP:Graphical Models and Image Processing, 55(1):1–19, Jan 1993.Google Scholar
  9. [9]
    J.Besag. On the statistical analysis of dirty pictures. J. Royal Statist. Soc., Series B, 48:259–302, 1986.Google Scholar
  10. [10]
    J.M.Bernardo and A.F.M.Smith. Bayesian Theory. Wiley, 1994.Google Scholar
  11. [11]
    L.Tierny. Markov Chains for exploring posterior distributions. Annals of Statistics, 22(5): 1701–1762, 1994.Google Scholar
  12. [12]
    P.J.Green. Reversible jump Markov Chain Monte Carlo computation and Bayesian model determination. Biometrika, 82(4):711–732, 1996.Google Scholar
  13. [13]
    S.Geman and D.Geman. Stochastic Relaxation, Gibbs Distributions and the Bayesian Restoration of Images. IEEE Trans. Patt. Anal & Machine Intell., 6(6):721–741, Nov 1984.Google Scholar
  14. [14]
    S.Richardson and P.J.Green. On Bayesian analysis of mixtures with an unknown number of components.-, Feb 1996.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1997

Authors and Affiliations

  • S. A. Barker
    • 1
  • P. J. W. Rayner
    • 1
  1. 1.Signal Processing and Communications GroupCambridge University Engineering Dept.CambridgeEngland

Personalised recommendations