This Article 
 Bibliographic References 
 Add to: 
Estimation of Generalized Multisensor Hidden Markov Chains and Unsupervised Image Segmentation
May 1997 (vol. 19 no. 5)
pp. 465-475

Abstract—This paper attacks the problem of generalized multisensor mixture estimation. A distribution mixture is said to be generalized when the exact nature of components is not known, but each of them belongs to a finite known set of families of distributions. Estimating such a mixture entails a supplementary difficulty: One must label, for each class and each sensor, the exact nature of the corresponding distribution. Such generalized mixtures have been studied assuming that the components lie in the Pearson system. Adaptations of classical algorithms, such as Expectation-Maximization, Stochastic Expectation-Maximization, or Iterative Conditional Estimation, can then be used to estimate such mixtures in the context of independent identically distributed data and hidden Markov random fields. We propose a more general procedure with applications to estimating generalized multisensor hidden Markov chains. Our proposed method is applied to the problem of unsupervised image segmentation. The method proposed allows one to: (i) identify the conditional distribution for each class and each sensor, (ii) estimate the unknown parameters in this distribution, (iii) estimate priors, and (iv) estimate the "true" class image.

[1] B. Benmiloud and W. Pieczynski, "Estimation des paramètres dans les chaînes de Markov cachées et segmentation d'images," Traitement du Signal, vol. 12, no. 5, pp. 433-454, 1995.
[2] J. Besag, "On the Statistical Analysis of Dirty Pictures," J. Royal Statistical Soc., Series B, vol. 48, pp. 259-302, 1986.
[3] B. Braathen, W. Pieczynski, and P. Masson, "Global and Local Methods of Unsupervised Bayesian Segmentation of Images," Machine Graphics and Vision, vol. 2, no. 1 pp. 39-52, 1993.
[4] B. Chalmond, "An Iterative Gibbsian Technique for Reconstruction of m-ary Images," Pattern Recognition, vol. 22, no. 6, pp. 747-761, 1989.
[5] R. Chellapa and A. Jain, eds., Markov Random Fields. Academic Press, 1993.
[6] Y. Delignon, A. Marzouki, and W. Pieczynski, Estimation of Generalized Mixture and Its Application in Image Segmentation IEEE Trans. Image Processing, vol. 6, no. 10, pp. 1364-1375, 1997.
[7] A.P. Dempster, N.M. Laird, and D.B. Rubin, "Maximum Likelihood From Incomplete Data via the EM Algorithm," J. Royal Statistical Soc., Series B, vol. 39, pp. 1-38, 1977.
[8] P.A. Devijver, "Hidden Markov Mesh Random Field Models in Image Analysis," Advances in Applied Statistics, Statistics and Images, vol. 1, pp. 187-227. Carfax Publishing Co., 1993.
[9] R.C. Dubes and A.K. Jain, "Random Field Models in Image Analysis," J. Applied Statistics, vol. 16, no. 2, pp. 131-164, 1989.
[10] G.D. Fornay, "The Viterbi Algorithm," Proc. IEEE, vol. 61, no. 3, pp. 268-277, 1973.
[11] S. Geman and G. Geman, "Stochastic Relaxation, Gibbs Distributions and the Bayesian Restoration of Images," IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 6, pp. 721-741, 1984.
[12] N. Giordana, "Segmentation non supervisée d'images multispectrales par chaînes de Markov cachées," PhD Thesis, Universitéde Technologie de Compiègne, 1996.
[13] R. Haralick and J. Hyonam, "A Context Classifier," IEEE Trans. GRS, vol. 24, pp. 997-1,007, 1986.
[14] N.L. Johnson and S. Kotz, Distributions in Statistics: Continuous Univariate Distributions, vol. 1. J. Wiley, 1970.
[15] R. Barta and M.W. Schranz, "Jessica—An Object-Oriented Hypermedia Publishing Processor," Computer Networks and ISDN Systems, vol. 30, no. 1-7, Apr. 1998, p. 281.
[16] S. Lakshmanan and H. Derin, “Simultaneous Parameter Estimation and Segmentation of Gibbs Random Fields Using Simulated Annealing,” IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 11, pp. 799-813, 1989.
[17] A.M. Law and W.D. Kelton, Simulation Modeling and Analysis, 2nd ed. McGraw Hill, 1991.
[18] A. Maffet and C. Wackerman, "The Modified Beta Density Function as a Model for Synthetic Aperture Radar Clutter Statistics," IEEE Trans. GRS, vol. 29, no. 2, 1991.
[19] J. Marroquin, S. Mitter, and T. Poggio, "Probabilistic Solution of Ill-Posed Problems in Computational Vision," J. Am. Statistical Assoc., vol. 82, pp. 76-89, 1987.
[20] P. Masson and W. Pieczynski, "SEM Algorithm and Unsupervised Segmentation of Satellite Images," IEEE Trans. GRS, vol. 31, no. 3, pp. 618-633, 1993.
[21] E. Mohn, N. Hjort, and G. Storvik, "A Simulation Study of Some Contextual Classification Methods for Remotely Sensed Data," IEEE Trans. GRS, vol. 25, pp. 796-804, 1987.
[22] A. Peng and W. Pieczynski, "Adaptive Mixture Estimation and Unsupervised Local Bayesian Image Segmentation, Graphical Models, and Image Processing," vol. 57, no. 5, pp. 389-399, 1995.
[23] W. Pieczynski, "Mixture of Distributions, Markov Random Fields, and Unsupervised Bayesian Segmentation of Images," Technical Report 122, L.S.T.A., UniversitéParis 6, 1990.
[24] W. Pieczynski, "Statistical Image Segmentation," Machine Graphics and Vision, vol. 1, nos. 1and 2, pp. 261-268, 1992.
[25] W. Qian and D.M. Titterington, "On the Use of Gibbs Markov Chain Models in the Analysis of Images Based on Second-Order Pairwise Interactive Distributions," J. Applied Statistics, vol. 6 no. 2, pp. 267-282, 1989.
[26] H.C. Quelle, Y. Delignon, and A. Marzouki, "Unsupervised Bayesian Segmentation of SAR Images Using the Pearson System," Proc. IGARSS '93, pp. 1,538-1,540, 1993.
[27] L.R. Rabiner, “Tutorial on Hidden Markov Model and Selected Applications in Speech Recognition,” Proc. IEEE, vol. 77, no. 2, pp. 257-285, 1989.
[28] R.A. Redner and H.F. Walker, "Mixture Densities, Maximum Likelihood and the EM Algorithm," SIAM Rev., vol. 26, pp. 195-239, 1984.
[29] W. Skarbek, Generalized Hilbert Scan in Image Printing, Theoretical Foundation of Computer Vision, R. Klette and W.G. Kropetsh, eds., pp. 45-57. Akademik Verlag, 1992.
[30] J. Tilton, S. Vardeman, and P. Swain, "Estimation of Context for Statistical Classification of Multispectral Image Data," IEEE Trans. GRS, vol. 20, pp. 445-452, 1982.
[31] L. Younes, "Parametric Inference for Imperfectly Observed Gibbsian Fields," Probability Theory and Related Fields, vol. 82, pp. 625-645, 1989.
[32] J. Zhang,“The mean field theory in EM procedures for blind Markov random field image restoration,” IEEE Transactions on Image Proc., vol. 2, no. 1, pp. 27-40, Jan. 1993.

Index Terms:
Multisensor data, mixture estimation, generalized mixture estimation, hidden Markov chain, Bayesian segmentation, unsupervised segmentation.
Nathalie Giordana, Wojciech Pieczynski, "Estimation of Generalized Multisensor Hidden Markov Chains and Unsupervised Image Segmentation," IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 19, no. 5, pp. 465-475, May 1997, doi:10.1109/34.589206
Usage of this product signifies your acceptance of the Terms of Use.