评价滤波算法的SAR图像模拟

SAR Image Simulation for Checking Filtering Algorithms

提出了一种简单的基于统计模型的SAR(合成孔径雷达)图像的模拟的算法.该算法首先根据自然场景(如海面、场地和森林等)的统计模型产生相关的Gamma分布,得到场景的后向散射系数图;然后根据乘法模型引入斑点噪声,得到复数的后向散射系数图;再将点面目标的后向散射系数叠加在后向散射系数图上;最后将该复数图像通过一个由SAR系统的系统传输函数定义的滤波器生成模拟的SAR图像.计算机模拟结果表明,该方法与以前的模拟算法比较,不仅能真实地反映SAR图像的特征,而且计算简单,为检验滤波算法提供了有力的工具.

A new simple method simulating synthetic aperture radar image based on statistic model is proposed in this paper. At the first a correlated gamma-distributed RCS variable is generated according to the statistic model of natural scene. The second stage, corresponding to introducing speckle in the image process according to product model, is achieved by generating a complex image, and adding to point and area target's RCS to the complex image. At last, the simulated SAR image is got by passing the complex image through a filter defined by the SAR system transform function. The result shows that this method is suitable to simulate SAR image, and the computation is very simple. It is a powerful tool for checking the filtering algorithms.