Joint Dimensionality Reduction, Classification and Segmentation of Hyperspectral Images

Dimensionality reduction, spectral classification and segmentation are the three main problems in hyperspectral image analysis. In this paper we propose a Bayesian estimation approach which gives a solution for these three problems jointly. The data reduction problem is modeled as a blind sources separation (BSS) where the sources are the images which must be mutually independent and piecewise homogeneous. To insure these properties, we propose hierarchical model for the sources with a common hidden classification variable winch-is modelled as a Potts-Markov field. The joint Bayesian estimation of this hidden variable as well as the sources and the mixing matrix of the BSS problem gives a solution for all the three problems of dimensionality reduction, spectra classification and segmentation of hyperspectral images. For the Bayesian computation, we propose to use either Gibbs sampling (GS) or mean field approximation (MFA) methods. A few simulation results illustrate the performances of the proposed method and some comparison with other classical methods of PCA and ICA used for BSS.