SCIA ’ 2001 Bergen , Norway , June 11 - 14 , 2001 1 COMPOUND DISTRIBUTIONS FOR RADAR IMAGES

In radar imagery, the statistical modeling is derived from random walk theory because the backscattered field is the sum of the contributions of the different scatterers at the observed surface. In this paper we describe how the dependence of the statistical properties of microwave returns from the ground can be built into a theoretical model in a consistent way. When the number of scatterers varies, the distribution of the resulting field may be non-Gaussian. Here we derive the field distribution based on the assumption of compound Poisson distributions of the number of scatterers. We obtain a new set of backscattered field distributions including well-known K and Rayleigh distributions. These novel classes of probability distributions are found to correlate well with distributions of radar clutter returns. We assess the relevance of the modeling with experimental radar data and we compare the fitting results of the new distributions to the K and Rayleigh ones. We finally point out t he explicit dependence between the statistical properties of the number of scatterers and the surface illuminated by the radar.