This Article 
 Bibliographic References 
 Add to: 
Unsupervised Texture Segmentation Using Markov Random Field Models
May 1991 (vol. 13 no. 5)
pp. 478-482

The problem of unsupervised segmentation of textured images is considered. The only explicit assumption made is that the intensity data can be modeled by a Gauss Markov random field (GMRF). The image is divided into a number of nonoverlapping regions and the GMRF parameters are computed from each of these regions. A simple clustering method is used to merge these regions. The parameters of the model estimated from the clustered segments are then used in two different schemes, one being all approximation to the maximum a posterior estimate of the labels and the other minimizing the percentage misclassification error. The proposed approach is contrasted with the algorithm of S. Lakshamanan and H. Derin (1989), which uses a simultaneous parameter estimation and segmentation scheme. The results of the adaptive segmentation algorithm of Lakshamanan and Derin are compared with a simple nearest-neighbor classification scheme to show that if enough information is available, simple techniques could be used as alternatives to computationally expensive schemes.

[1] S. Lakshmanan and H. Derin, "Simultaneous parameter estimation and segmentation of Gibbs random fields using simulated annealing,"IEEE Trans. Pattern Anal. Machine Intell., vol. 11, pp. 799-813, Aug. 1989.
[2] H. Derin and H. Elliott, "Modeling and segmentation of noisy and textured images using Gibbs random fields,"IEEE Trans. Pattern Anal. Machine Intell., vol. PAMI-9, pp. 39-55, Jan. 1987.
[3] B. S. Manjunath, T. Simchony, and R. Chellappa, "Stochastic and deterministic networks for texture segmentation,"IEEE Trans. Acoust., Speech, Signal Processing, vol. 38, pp. 1039-1049, June 1990.
[4] D. Geman, S. Geman, C. Graffigne, and P. Dong, "Boundary detection by constrained optimization,"IEEE Trans. Pattern Anal. Machine Intell., vol. 12, pp. 609-628, July 1990.
[5] F. S. Cohen and Z. Fan, "Unsupervised textured image segmentation," Dep. Elec. Eng., Univ. Rhode Island, Tech. Rep. 86-1, June 1986.
[6] F. S. Cohen and D. B. Cooper, "Simple parallel hierarchical and relaxation algorithms for segmenting noncausal Markovian fields,"IEEE Trans. Pattern Anal. Machine Intell., vol. PAMI-9, pp. 195-219, March 1987.
[7] R. Chellappa, "Two-dimensional discrete Gaussian Markov random field models for image processing," inProgress in Pattern Recognition 2, L. N. Kanal and A. Rosenfeld, Eds. Amsterdam, The Netherlands: Elsevier Science (North-Holland), 1985, pp. 79-112.
[8] R. Kashyap and R. Chellapa, "Estimation and choice of neighbors in spatial interaction models of images,"IEEE Trans. Inform. Theory, vol. IT-29, pp. 60-72, Jan. 1983.
[9] J. Besag, "On the statistical analysis of dirty pictures,"J. Royal Statist. Soc. B, vol. 48, pp. 259-302, 1986.
[10] J. L. Marroquin, "Probabilistic solution of inverse problems," Ph.D. dissertation, Artificial Intell. Lab., Massachusetts Inst. Technol., Cambridge, Sept. 1985.
[11] C. S. Won and H. Derin, "Unsupervised image segmentation using Markov random fields--Part I: Noisy images," preprint.

Index Terms:
unsupervised texture segmentation; picture processing; pattern recognition; Gauss Markov random field; Markov processes; parameter estimation; pattern recognition; picture processing
B.S. Manjunath, R. Chellappa, "Unsupervised Texture Segmentation Using Markov Random Field Models," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 13, no. 5, pp. 478-482, May 1991, doi:10.1109/34.134046
Usage of this product signifies your acceptance of the Terms of Use.