基于单尺度脊波变换的阈值滤波方法  

Image threshold denoising algorithm based on monoscale ridgelet transform

作　　者：谭兮[1] 于会军[1] 

机构地区：[1]湖南工业大学电气工程系,湖南株洲412008

出　　处：《计算机应用》2007年第4期966-969,共4页journal of Computer Applications

摘　　要：分析了高斯白噪声在单尺度脊波域中的统计性质,提出了一种新的基于单尺度脊波变换的阈值滤波算法。仿真结果表明,这种算法不仅比传统基于小波变换的各种滤波算法有更高的PSNR值,而且能更好地保持图像细节。The statistical property of Gaussian White noise in the Monoscale Ridgelet domain was analyzed, and a new image threshold denoising algorithm based on monoseale ridgelet transform was proposed. The experimental results demonstrate that the new denoising algorithm substantially surpasses those based on wavelet transforms, both visually and in terms of PSNR.

关 键 词：滤波 脊波变换 单尺度脊波变换 小波变换 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]