POLSAR image factorization and its extended applications

Abstract

POLSAR image factorization is proposed as an extension of polarimetric SAR incoherent target decomposition. It simultaneously estimates a dictionary of meaningful atom scatterers and their corresponding spatial distribution maps from POLSAR image. Both model-based and eigenanalysis-based decompositions can be seen as special cases of image factorization under specific constraints. It can be solved via non-negative matrix factorization (NMF). Based on the obtained distribution maps, extended applications such as speckle reduction and image classification can be conducted.