Hierarchical Markovian segmentation of multispectral images for the reconstruction of water depth maps J.-N. Provost, a,1 C. Collet, a, * P. Rostaing, a P. P erez, b and P. Bouthemy b a IRENav/GTS-BP 600-29240, Brest Naval, France b IRISA/INRIA Rennes, Campus Universitaire de Beaulieu, 35042 Rennes Cedex, France Received 15 January 2002; accepted 28 July 2003 Abstract This paper presents an unsupervised method to segment multispectral images, involving a correlated non-Gaussian noise. The efficiency of the Markovian quadtree-based method we propose will be illustrated on a satellite image segmentation task with multispectral observa- tions, in order to update nautical charts. The proposed method relies on a hierarchical Mar- kovian modeling and includes the estimation of all involved parameters. The parameters of the prior model are automatically calibrated while the estimation of the noise parameters is solved by identifying generalized distribution mixtures [P. Rostaing, J.-N. Provost, C. Collet, Proc. International Workshop EMMCVPRÕ99: Energy Minimisation Methods in Computer Vision and Pattern Recognition, Springer Verlag, New York, 1999, p. 141], by means of an iterative conditional estimation (ICE) procedure. Generalized Gaussian (GG) distributions are consid- ered to model various intensity distributions of the multispectral images. They are indeed well suited to a large variety of correlated multispectral data. Our segmentation method is applied to Satellite Pour lÕObservation de la Terre (SPOT) remote multispectral images. Within each segmented region, a bathymetric inversion model is then estimated to recover the water depth map. Experiments on different real images have demonstrated the efficiency of the whole process and the accuracy of the obtained results has been assessed using ground truth data. * Corresponding author. Present address: Universite Louis Pasteur, ENSPS-LSIIT UMR CNRS 7005, Bd S. Brant, 67400 Illkirch, France. Fax: +33-3-90-24-43-42. E-mail addresses: Jean-Noel.Provost@insa-rennes.fr (J.-N. Provost), christophe.collet@ensps. u-strasbg.fr (C. Collet), bouthemy@irisa.fr (P. Bouthemy). 1 Present address: IETR Groupe Image, UMR CNRS 6164, INSA-Rennes, CS 14315, 35043 Rennes Cedex, France. 1077-3142/$ - see front matter Ó 2003 Elsevier Inc. All rights reserved. doi:10.1016/j.cviu.2003.07.004 Computer Vision and Image Understanding 93 (2004) 155–174 www.elsevier.com/locate/cviu