将光学图像加性噪声抑制的非线性扩散方程引入到SAR图像相干斑抑制中，通过分析SAR图像的相干斑噪声模型在功率图像域以及功率图像的对数域上的表现特性，说明了基于加性噪声模型的非线性扩散方程在SAR图像噪声抑制中的应用可能性，并由此设计了基于非线性扩散方程的SAR图像噪声抑制算法。计算结果表明，与现有的SAR噪声抑制方法相比，非线性扩散方程方法具有更强的对背景区域的噪声抑制能力，且能更好地保持目标特征。

This paper takes advantge of the anisotropic diffusion for additive noise reducing to reduce the speckle in SAR image. It analyses the distribution of speckle model of SAR image and its Log image, and the result shows that using the anisotropic diffusion for additive noise reducing to reduce the speckle is feasible. Then it gives an algorithm for speckle reducing based on anisotropic diffusion. Numerical results show that the new model has better denoising result than some known methods, and can heep target feature better.

TN959.73

国家自然科学基金资助项目（No.60272013）;全国优秀博士论文作者专项基金资助项目（No.200140）.