Abstract
This paper is concerned with hierarchical Markov Random Field (MRF) models and their applications to multispectral image segmentation. We present an extension of the classic Gaussian model for the modelization of the data likelihood based on a Generalized Gaussian (GG) model, requiring a “shape parameter”. In order to obtain an unsupervised multispectral image segmentation, we develop a two step algorithm. In the first step, we estimate the parameters associated with a causal Markovian model (on a quad-tree1) and a generalized Gaussian modeling for the data-driven term, by using an Iterative Conditional Estimation (ICE algorithm [16]). One of the originality of this paper consists in explicitly decorrelate the multispectral observations during the estimation step on a quad-tree structure. A second step gives the segmentation map obtained with the estimated parameters, according to the Modes of Posterior Marginals (MPM) estimator. The main motivation of the paper is to extend the variety of noise models which results of the distribution mixture on multispectral images. Some results on synthetic and SPOT images validate our approach.
Acknowledgments: The authors thank P. Pérez and P. Bouthemy for fruitful comments and suggestions.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
X. Descombes A. Lorette and J. Zérubia. Extraction desz onesurba inesf ondée sur une analyse de la texture par modélisation markovienne. Rapport interne, INRIA-Sophia-Antipolis, 1998.
V. R. Algazi and R. M. Lerner. Binary detection in white non-Gaussian noise. Technical Report DS-2138, Lab., Lexington, MA, 1964.
G. Celeux and J. Diebolt. L’algorithme SEM: un algorithme d’apprentissage probabiliste pour la reconnaissance de mélange de densités. Revue de statistiques appliquées, 34(2):35–52, 1986.
G. Celeux and J. Diebolt. A random imputation principle: the stochastic EM algorithm. INRIA research report, (901):1–19, septembre 1988.
R. Chellappa and R. L. Kashyap. Statistical inference in gaussian markov random field models. IEEE Computer Society Conference on Pattern Recognition and Image Processing, pages77–80, juin 1982.
Y. Dong, B. Forster, and A. Milne. Segmentation of radar imagery using gaussian markov random field models and wavelet transform techniques. IEEE International Geoscience and Remote Sensing Symposium Proceedings (IGARSS-97), 4:2054–2056, 1997.
R. J. Elliott and L. Aggoun. Estimation for discret Markov random fields observed in gaussian noise. IEEE Transaction on Information Theory, IT-40(5):739–742, September 1994.
N. Giordana and W. Pieczynski. Estimation of generalized multisensor hidden Markov chains and unsupervised image segmentation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 19(5):465–475, 1997.
C. Graffigne, F. Heitz, P. Pérez, F. Prêteux, M. Sigelle, and J. Zerubia. Hierarchical Markov random field modelsap plied to image analysis: a review. In SPIE Neural Morphological and Stochastic Methods in Image and Signal Processing, volume 2568, pages2–17, San Diego, 10-11 July 1995.
P. Pérez J.-M. Laferté, F. Heitz and E. Fabre. Hierarchical statistical models for the fusion of multiresolution image data. In Proc. Int. Conf. Computer Vision, pages908–913, Cambridge, USA, June 1995.
M. Kanefsky and J. B. Thomas. On polarity detection schemes with non-Gaussian inputs. J. Franklin Inst., 280:120–138, 1965.
J.-M. Laferté, P. Pérez, and F. Heitz. Discrete markov image modeling and inference on the quad-tree. Technical Report 1198, INRIA, July 1998.
M. Mignotte, C. Collet, P. Pérez, and P. Bouthemy. Unsupervised segmentation applied on sonar images. In Proc. International Workshop EMMCVPR’97: Energy Minimisation Methods in Computer Vision and Pattern Recognition (Springer editor), volume 1223, pages 491–506, Venice, Italy, May 1997.
M. Mignotte, C. Collet, P. Pérez, and P. Bouthémy. Three-class markovian segmentation of high resolution sonar images. Journal of Computer Vision and Image Understanding, submitted.
T. T. Pham and R. J. P. deFigueiredo. Maximum likelihood estimation of a class of non-gaussian densities with application to l p deconvolution. IEEE Trans. on Acoustics, Speech and Signal Processing, 37(1):73–82, january 1989.
W. Pieczynski. Champs de Markov cachéset estimation conditionnelle itérative. Traitement du signal, 11(2):141–153, 1994.
W. Pieczynsky, J. Bouvrais, and C. Michel. Unsupervised bayesian fusion of correlated sensors. In First International Conference on Multisource-Multisensor Information Fusion, Las Vega, Nevada, USA, June 6-9 1998.
J. A. Provine and R. M. Rangayyan. Lossless compression of Peano scanned images. Journal of Electronic Imaging, 3(2):176–181, 1994.
R. Y. Rubinstein. Simulation and Monte Carlo Method. NY: Wiley, 1987. 155
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1999 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Rostaing, P., Provost, J.N., Collet, C. (1999). Unsupervised Multispectral Image Segmentation Using Generalized Gaussian Noise Model. In: Hancock, E.R., Pelillo, M. (eds) Energy Minimization Methods in Computer Vision and Pattern Recognition. EMMCVPR 1999. Lecture Notes in Computer Science, vol 1654. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-48432-9_11
Download citation
DOI: https://doi.org/10.1007/3-540-48432-9_11
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-66294-5
Online ISBN: 978-3-540-48432-5
eBook Packages: Springer Book Archive