一种基于K邻域平均法的小波图像恢复算法  

Algorithm of Wavelet Image Recovery Based on K Neighborhood Average Method

作　　者：赵晨萍[1] 郭运瑞[1] 李登峰[2,3] 宋锦萍[2,3] 

机构地区：[1]河南科技学院,河南新乡453003 [2]河南大学应用数学研究所,河南开封475001 [3]河南大学数学与信息科学学院,河南开封475001

出　　处：《现代电子技术》2009年第2期148-150,共3页Modern Electronics Technique

基　　金：国家自然科学委员会与中国工程物理研究院联合基金资助项目(10576013);河南省自然科学基金资助项目(0611053200);河南省教育厅基础研究资助项目(2006110003);河南科技学院校选项目基金资助项目(0655)

摘　　要：小波图像压缩及图像消噪算法近年来得到了广泛的重视和发展,其处理的核心部分就是图像的边缘,各种算法都会产生图像边缘的锯齿状震荡,即Gibbs现象。针对小波图像压缩中的振荡现象,首先应用Kirsch算子检测图像边缘,然后利用区域内像素灰度的相关性及小波系数的相关性,提出一种基于小波变换的K邻域平均法,最后通过仿真及信噪比的比较说明了此算法的有效性和可行性。The algorithms of image compression and image denoising using wavelet have been widely concerned and studied. How to keep the image edges is the important problem in all the algorithms. However,the Gibbs phenomenon can be produced along with all kinds of image procession. In order to eliminate the oscillations and improve the reconstruction image quality,the article firstly detects edge of the image by using Kirsch operators, then presents the K neighborhood average algorithm-based according to dependence of the image pixels and the wavelet coefficients,and finally shows the feasible and efficient performance by simulation experiments and signal - to - noise ratio.

关 键 词：小波变换 Gibbs现象 KIRSCH算子 K邻域平均法 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]