Hypothesis Testing for Texture Discrimination using the Geodesic Distance in SAR imagery under the G0I Model

NARANJO TORRES, JOSE; GAMBINI, JULIANA; LUCINI, MARÍA MAGDALENA; FRERY, ALEJANDRO CESAR

Valdivia

Conferencia; XIX Conference on Nonequilibrium Statistical Mechanics and nonlinear Physics,; 2016

In this work we analyze a SAR (Synthetic Aperture Radar) region discrimination method based on the Geodesic Distance (GD) and the estimation of the parameters of a statisticalmodel largely used for this kind of data. The statistical model used is the $\mathcal{G}_I^0$distribution which is indexed by three parameters: the number oflooks ($L$), a scale parameter ($\gamma$), and a parameter ($\alpha$) related to the roughness ortexture of the backscatter. This fact is one of the reasons why the estimation of $\alpha$receives a great deal of attention in the literature.This paper presents a new method to measure the separability between regions in SAR imagery using the GD (presented by Rao), under the $\mathcal{G}_I^0$ distribution. In order to asses the performance of the texture discrimination method, a hypothesis test is used. We derived closed form for the GD between models that describe several practical situations, assuming the number of looks known, for same and different texture and for same and different scale.The parameters, in each case, are estimated using the Maximum Likelihood method because we are specially interested in its asymptotic properties.