Estimate of Distribution for SAR Image Based on the Generalized Rayleigh Distribution

Generalized Rayleigh distribution was deduced for the amplitude SAR image according to the generalized central limit theorem, and the parameters were estimated by the negative order moment. Based on the disadvantage of single-point estimation, global search was suggested to achieve the optimal parameters. Owing to lack of analytical expression of the probability density function, the generalized Rayleigh distribution was calculated by the series expansion, and its convergence was analyzed as well as some cases of divergence. Estimate of distribution for SAR image demonstrates that the estimate of distribution based on the generalized Rayleigh distribution is superior with lower mean square error (MSE) compared to the one based on the traditional Rayleigh distribution, and the global search for the optimal parameters reduces the probability of divergence significantly.