Abstract: This paper reviews probability distribution models in the study of statistical modeling of Synthetic Aperture Radar（SAR） imagery. According to the origin of each model, all models can be divided into two categories：the statistical models of prior assumptions and the empirical distribution models. The Method of Logarithmic Cumulants（MoLC） which is a new method for parameter estimation is introduced；furthermore, the estimated expressions of logarithmic cumulants of statistical models are computed based on the MoLC. Theevaluation criterions of modeling precision are developed and then implemented in the experiment of modeling real SAR images. Finally, the suitable terrain of SAR imagery for each statistical model is concluded from experimental results.

Key words: Synthetic Aperture Radar（SAR） imagery, probability distribution model, parameter estimation, precision evaluation

摘要： 综述了合成孔径雷达（SAR）图像统计建模研究中的概率分布模型，按照模型起源将所有模型分为了先验假设统计模型和经验分布模型两大类。介绍了一种模型参数估计的新方法——对数累量法，并根据对数累量法计算了统计模型参数估计的对数累量表达式；发展了统计模型建模精度的评估准则，并应用到了SAR图像数据统计建模实验中。通过实验结果得出了每一种统计模型适宜建模的SAR图像地表类型。

关键词: SAR图像, 概率分布模型, 参数估计, 精度评估