MODELING SAR IMAGES BASED ON A GENERALIZED GAMMA DISTRIBUTION FOR TEXTURE COMPONENT

In the applications of synthetic aperture radar (SAR) data, a crucial problem is to develop precise models for the statistics of the pixel amplitudes or intensities. In this paper, a new statistical model, called simply here Gii, is proposed based on the product model by assuming the radar cross section (RCS) components (texture components) of the return obey a recently empirical generalized Gamma distribution. Meanwhile, we demonstrate theoretically that the proposed Gii model has the well-known K and G - distributions as special cases. We also derived analytically the estimators of the presented Gii model by applying the \method-of-log-cumulants" (MoLC). Finally, the performance of the proposed model is tested by using some measured SAR images.