Shearlet域变换与复扩散方程结合的SAR图像降斑

SAR image despeckling based on edge detection and the plural pervasion equation in the Shearlet domain

针对大多数降斑算法不能很好地在实现区域平滑的同时保持图像细节边缘这一问题,提出了一种新的复扩散方程,结合Shearlet变换给出了一种模极大值边缘检测方法.将获得图像边缘与降斑平滑的结果相结合,在有效实现降斑平滑的同时保持了图像的边缘.实验结果表明,该方法对SAR图像中的点目标、边缘和纹理具有较好的保持能力,在有效降斑的同时,更多地保留了图像的边缘和细节.

It is difficult for many current despeckling methods to produce good region smoothing and edge preservation simultaneously.In this paper,a new Plural Pervasion equation is designed,which combines a Shearlet transform and edge detection of module maximum,and a new SAR image despeckling method is proposed.Experimental results indicate that the method possesses advantages of speckle reduction,and point target and edge preservation.