Real-Time Restoration and Segmentation Algorithms for Hidden Markov Mesh Random Fields Image Models

  • Pierre A. Devijver
  • Michel M. Dekesel
Part of the NATO ASI Series book series (volume 42)


This paper addresses the image restoration and segmentation problems under the assumption that images can be represented by hidden Markov mesh random fields models. We outline coherent approaches to both the problems of image segmentation and restoration (pixel labeling) and model acquisition (learning). We exhibit a real-time labeling algorithm for a 3rd order Markov mesh which achieves minimal complexity. We develop a learning technique which permits to estimate the model parameters without ground truth information. We display experimental results which demonstrate that the approach is subjectively relevant to the image restoration and segmentation problems.


Markov Random Field Image Restoration Segmentation Problem Minimal Complexity Decision Directed 
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]
    K. Abend, T.J. Harley, and L.N. Kanal, “Classification of binary random patterns,” IEEE Trans. Inform. Theory, IT-11, pp. 538–544, Oct. 1965.CrossRefMathSciNetGoogle Scholar
  2. [2]
    J. Besag, “On the statistical analysis of dirty pictures,” paper read at the SERC Research Workshop on Statistics and Pattern Recognition, Edinburgh, July 1985.Google Scholar
  3. [3]
    H. Derin, H. Elliot, R. Christi, and D. Geman, “Bayes smoothing algorithms for segmentation of binary images modeled by Markov random fields,” IEEE Trans. Pattern Anal., Machine Intell., PAMI-6, pp. 707–720, Nov. 1984.CrossRefGoogle Scholar
  4. [4]
    H. Derin, and H. Elliot, “Modeling and segmentation of noisy and textured images using Gibbs random fields,” IEEE Trans. Pattern Anal., Machine Intell., PAMI-9, pp. 39–55, 1987.CrossRefGoogle Scholar
  5. [5]
    P.A. Devijver, “Probabilistic labeling in a hidden second order Markov mesh,” in Pattern Recognition in Practice II, E. Gelsema, and L.N. Kanal Eds., Amsterdam: North Holland, 1985, pp. 113–123.Google Scholar
  6. [6]
    P.A. Devijver, “Segmentation of binary images using third order Markov mesh image models,” in Proc. 8th Internat. Conf. on Pattern Recognition, Paris, Oct. 1986, pp. 259–261.Google Scholar
  7. [7]
    P.A. Devijver, and M.M. Dekesel, “Learning the parameters of a hidden Markov random field image model: A simple example,” in Pattern Recognition Theory and Applications, P. Devijver and J. Kittler Eds., Heidelberg: Springer, 1987, pp. 141–163.Google Scholar
  8. [8]
    P.A. Devijver and M.M. Dekesel, “Cluster analysis under Markovian dependence with application to image segmentation,” to appear in Proc. 1st Conf. Intern. Fed. Classification Societies, Aachen, June 1987.Google Scholar
  9. [9]
    P.A. Devijver and M.M. Dekesel, “Algorithmes d’apprentissage de modèles Markoviens d’images,” to appear in Proc. 6ème Congres RFIA, Antibes, Nov. 1987.Google Scholar
  10. [10]
    S. Geman, and D. Geman, “Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images,” IEEE Trans. Pattern Anal., Machine Intell., PAMI-6, pp. 721–741, Nov. 1984.CrossRefGoogle Scholar
  11. [11]
    D. Geman, S. Geman, and C. Graffigne, “Locating texture and object boundaries,” in Pattern Recognition Theory and Applications, P.A. Devijver and J. Kittler Eds., Heidelberg: Springer-Verlag, 1987.Google Scholar
  12. [12]
    J. Haslett, “Maximum likelihood discriminant analysis on the plane using a Markovian model of spatial context,” Pattern Recognition, 18, pp. 287–296, 1985.CrossRefMATHGoogle Scholar
  13. [13]
    F.-C. Jeng, and J.W. Woods, “On the relationship of the Markov mesh to the NSHP Markov chain,” Pattern Recognition Letters, 5, pp. 273–279, 1987.CrossRefMATHGoogle Scholar
  14. [14]
    L.N. Kanal, “Markov mesh models,” Computer Graphics and Image Processing, 12, pp. 371–375, 1980 (also in Image Modeling, A. Rosenfeld Ed., New York: Academic Press, 1981, pp. 239-243).Google Scholar
  15. [15]
    R. Kinderman and J.L. Snell, Markov Random Fields and their Applications, Providence Rl: American Mathematical Society, 1980.Google Scholar
  16. [16]
    L.A. Liporace, “Maximum likelihood estimation for multivariate observations of Markov sources” IEEE Trans. Inform. Theory, IT-28, 729–734, 1982.CrossRefMathSciNetGoogle Scholar
  17. [17]
    D.K. Pickard, “A curious binary lattice process,” Journal Applied Probability, 14, pp. 717–731, 1977.CrossRefMATHMathSciNetGoogle Scholar
  18. [18]
    D.K. Pickard, “Unilateral Markov fields,” Adv. Applied Probability, 12, pp. 655–671, 1980.CrossRefMATHMathSciNetGoogle Scholar
  19. [19]
    R.A. Redner, and H.F. Walker, “Mixture densities, maximum likelihood and the EM algorithm,” Siam Review, 26, pp. 195–239, 1984.CrossRefMATHMathSciNetGoogle Scholar
  20. [20]
    S. Yakovitz, “Unsupervised learning and the identification of finite mixtures,” IEEE Trans. Inform. Theory, IT-16, pp. 330–338, 1970.CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1988

Authors and Affiliations

  • Pierre A. Devijver
    • 1
  • Michel M. Dekesel
    • 1
  1. 1.Philips Research LaboratoryBrusselsBelgium

Personalised recommendations