增强型Shearlet域SAR图像去噪

针对合成孔径雷达(SAR)图像固有的相干斑噪声,提出了一种增强型Shearlet域SAR图像滤波去噪算法。该算法首先分析SAR图像的局部特性,利用局部特性,将图像分成均质区域、非均质区域和边缘区域。对均质区域,用均值滤波器进行滤波去噪;对非均质区域,用Shearlet变换处理去噪;对边缘区域,进行直接保留。实验结果证明,该算法不仅能够显著消除相干斑噪声,明显的提高视觉效果,而且能够很好的保持边缘细节和纹理信息。

SAR image de-noising in the enhanced Shearlet domain

Aiming at the speckle noise of synthetic aperture radar （SAR） image,an enhanced SAR image filtering de-noising algorithm in Shearlet transform domain is proposed.According to the analysis of the local properties of SAR image,taking advantage of the local properties,the SAR image will be divided into homogeneous area,nonhomogeneous area and edge area.Then,average filtering method is applied to the homogeneous area.Shearlet transform de-nosing method is applied to the non-homogeneous area and edge area retained directly.The experimental result shows that this algorithm not only can effectively eliminate the noise,improve the visual effect of the image,but also maintain edge details and texture information better.