基于Curvelet变换的自适应阈值图像去噪方法  

Adaptive threshold image denoising methed based on curvelet transform

作　　者：王海珍[1] 吴爱弟[1] 

机构地区：[1]天津工程师范学院理学院

出　　处：《天津理工大学学报》2010年第1期38-40,共3页Journal of Tianjin University of Technology

基　　金：天津市自然科学基金(08JCYBJC12100)

摘　　要：与小波变换相比,Curvelet变换能更好地表达图像的边缘和细节,因此更适合做图像处理.提出了一种基于第二代Curvelet变换的自适应阈值图像去噪方法,采用不同的阈值自适应地对不同尺度和方向的Curvelet系数进行阈值处理.实验结果表明,提出的方法在去除噪声的同时,能更好地保留图像的细节.去噪后的图像有更高的峰值信噪比和更好的视觉效果.The curvelet transform represents edges and details of image better than wavelet transform, and is therefore well - suited for image processing. A adaptive threshold image denoising method based on the second curvelet transform is proposed. The curvelet transform coefficients in different scales and directions are filtered with adaptive threshold. Experiments show that the method proposed yields denoised image with higher quality recovery of edges. It is capable of achieving the higher peak signal - noise - ratio and giving better visual quality.

关 键 词：CURVELET变换 图像去噪 小波变换 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]