Abstract:

A method of simulating high-resolution synthetic aperture radar (SAR) clutter images is proposed. Firstly, the properties of two statistical distributions (generalized Gamma distribution and generalized compound distribution) used for modeling high-resolution SAR clutter images are analyzed. Then, by applying the nonlinear transformation method, the analytical expressions of generating functions corresponding to the two models are derived. After that, with parameters estimated from the real SAR images and the given ones, clutter images following the two distributions are simulated respectively. The experimental results show that the histograms of the simulated images fit the theoretical probability density functions well, which validates the effectiveness of the proposed method.