基于Contourlet域K-L变换的SAR图像相干斑抑制

针对SAR图像相干斑噪声去除问题,提出了一种基于多尺度分解的Contourlet域K-L变换的SAR图像去噪的新方法。方法首先对源图像进行Contourlet分解,在不同频段的子带图像中,利用K-L变换进行能量保持即提出信号的主要特征,用重构图像来进行去噪,最后通过Contourlet逆变换得到去噪之后的图像。在SAR图像上的实验结果表明,方法不仅较好地保持了图像的纹理和细节特征及边缘特征,且信噪比也较高。

SAR image denoising based on Contourlet domain K-L transform

For SAR image speckle noise reduction problems,a domain based on multi-scale decomposition Contourlet K-L transform new method of SAR image denoising is presented in this paper.Firstly,the source image Contourlet decomposed in different frequency sub-band image,using K-L transform the energy that is made to keep the main features of the signal with the reconstructed image to remove the noise,and finally through Contourlet after the inverse transform denoising image.SAR images in the experimental results show that the method is not only well maintained the details of the characteristics of image texture and edge features,and the signal to noise ratio is high.