基于模拟退火算法的SAR图像复原

Restoration of SAR images based on simulated annealing

根据吉布斯马尔可夫随机场模型和 SAR图像斑点噪声的伽玛分布统计特性 ,应用模拟退火算法 ,实现雷达截面的全局优化伪似然估计。本文将新解 4选 1的方法从极化 SAR图像推广到一般 SAR图像 ;提出一个实用有效的图像特征检测法 ;说明了一阶模型和零阶模型之间的联系 ,并给出满意的处理结果。

Based on Gibbs-Markov random field model and the Gamma distribution property of SAR image's speckle noise, global optimal pseudo likelihood estimation of radar cross-section can be achieved with simulated annealing algorithm. In this article, J. Schou's method is extended from polarimetric SAR images to usual images; a simple and effective method is presented for detecting structural information in SAR images; furthermore, the relationship between the first-order model and zero-order model is shown.