基于NSCT变换与自蛇扩散的SAR图像相干斑抑制算法

针对SAR图像中乘性相干斑噪声的抑制与边缘保护问题,提出了一种基于非下采样轮廓波变换(NSCT)与自蛇扩散的抑斑新算法。该算法先利用NSCT变换对SAR图像进行多层子带分解,然后借助自蛇扩散对SAR图像不同子带分别实施参数不同的扩散滤波,最后对各去噪子带进行NSCT重构获得的SAR图像再次进行自蛇扩散滤波处理,从而实现带有边缘保护与增强的SAR图像相干斑抑制。实验表明,与多种传统抑斑算法相比,本文算法在相干斑抑制与边缘保护性能上均有明显提升。

Despeckling Algorithm for SAR Images Based on NSCT and Self-snake Diffusion

Aiming at multiplicative speckle noise suppression and edge protection in SAR images, a new despeckling algorithm based on non-subsampled contourlet transform (NSCT) and self-snake diffusion is proposed. The SAR image is firstly decomposed into multilayer sub-bands by the NSCT. Then, different sub-bands are filtered by self-snake diffusions with different filtering parameters, respectively. The reconstructed SAR image is obtained by the inverse NSCT based on filtered sub-bands. In order to achieve well speckle reduction with edge protection and enhancement, the reconstructed SAR image is processed by self-snake diffusion again. The experimental results and comparisons with several traditional algorithms show that the proposed algorithm can significantly improve performances in speckle reduction and edge protection.