基于Contourlet变换的图像自适应阈值去噪方法

Adaptive Threshold-Denoising Method of Image Based on Contourlet Transform

论文研究了一种基于Contourlet变换的图像去噪方法。Contourlet变换是一种有效的图像表示方式,可以更好地表示图像中的边缘信息。根据Contourlet系数的不同特性,论文使用基于系数局部方差的自适应阈值改进了原有阈值,并结合循环平移Contourlet变换改进了阈值去噪算法。实验结果表明,与Contourlet软阈值、Contourlet硬阈值去噪相比,论文算法更好地保留了图像的轮廓细节,提高了图像的PSNR。

A class of image denoising method based contourlet transform is studied.Contourlet transform is an efficient method for image representation,it has better performance at portray the texture information of the image.According to the different characteristics of Contourlet coefficient,a denoising algorithm is studied based on adaptive threshold and the cyclespinning Contourlet transform.Experiments on image denoising shows that,compared to the Contourlet soft-threshold and Contourlet hard-threshold,the algorithm in this paper can keep more image detail and improve the PSNR.