Bayesian classification of multi-look polarimetric SAR images with a generalized multiplicative speckle model

In this paper, a maximum likelihood (ML) classification algorithm is proposed to classify multi-look polarimetric SAR images. This algorithm considers a generalized multiplicative speckle model in which three texture factors are assumed to separately affect three polarization channels. We derive the ML estimation of the texture parameters for each polarization channel with the complex Wishart distribution of the multi-look speckle covariance matrix, and design the corresponding ML classifier according to the Bayesian criterion. Both the texture class statistics and the discriminant function are given in simple closed forms. Further, a method for adaptively producing the a priori probabilities is also presented in order to improve the classification accuracy. This method utilizes the contextual information in a forward procedure, and does not need any iteration. With the NASA/JPL L-band 4-look polarimetric SAR data, the effectiveness of the presented classification algorithm is demonstrated, and using of the adaptive a prion probabilities is shown to result in improved classifications.