一种空域和频域相结合的图像消噪方法  

Novel method of image denoising combining spatial space and frequency domain

作　　者：宋瑞霞[1] 李亚楠[1] 张巧霞[1] 姚东星[1] 陈曦[1] 

机构地区：[1]北方工业大学理学院,北京100041

出　　处：《计算机工程与应用》2011年第34期184-186,211,共4页Computer Engineering and Applications

基　　金：国家自然科学基金(No.10631080;No.10771002);北京市自然科学基金(No.1102017);澳门科学技术发展基金(No.008/2008/A1)~~

摘　　要：提出一种基于非均匀剖分和V变换相结合的图像消噪新方法。非均匀剖分逼近在空域上视像素的灰度值为拟合数据,依据最小二乘法原理,对含噪图像进行非均匀三角剖分,使图像表示为一个分片多项式,它能保持图像的边缘及细节特征,并通过对剖分精度的控制来实现图像的消噪;V系统是一类L2[0,1]空间上的完备正交系,具有多小波的多分辨特性,利用相应的V变换将图像变换到频域,通过对高频低频的系数处理来达到消噪的目的。结合非均匀剖分逼近和V变换两者的优势,将两个方法的消噪结果加权平均,得到一种新的消噪方法。实验结果表明了该方法的消噪效果比很多经典方法更好。A new images denoising method based on non-uniform partition approximation and V-transform is proposed.The key point of non-uniform partition approximation is that the gray values of the pixels are regarded as fitting data,which are fitted with the least squares method,and then the digital image is expressed as a piecewise polynomial.It can retain the edge as well as detailed characteristics of image.The non-uniform partition approximation can realize image denoising by controlling partition precision value.V-system is a class of complete orthogonal function system on L 2 [0,1],with the characteristics of the multi-resolution of multi-wavelet.It can transform a digital image from spatial space to frequency domain,and by processing the coefficients in high frequency and low frequency to realize image denoising.This paper combines the advantages of these two methods to process image deniosing.The experimental results show that the proposed method is efficient to improve the quality of denoising images.

关 键 词：非均匀剖分 V系统 离散V变换 图像消噪 空域 频域 

分 类 号：TP31[自动化与计算机技术—计算机软件与理论]