Estimation of sea-ice SAR clutter statistics from Pearson's system of distributions

SAR images can be used to help ship routing in sea-ice conditions. In this study, we focus on the Antarctic region where no multi-year ice nor big ice floes are to be found. As a matter of fact, each clutter obeys to a backscattering mechanism that induces a specific pixel distribution and our attempt is to identify automatically the correct distribution for each ice type. The problem is that of generalized mixture estimation and unsupervised image classification. In this work, we modelled the mixture with distributions from the Pearson's system. Parameters estimation is realized according to the ICE algorithm in the context of hidden Markov chains. The results obtained from the Pearson's system are compared to the ones obtained with a classical mixture of Gaussian distributions.