一种基于小波变换的数据除噪方法  

Denosing method based on wavelet transform

作　　者：曾祥利[1] 傅彦[1] 青华平[1] 

机构地区：[1]电子科技大学计算机科学与工程学院,四川成都610054

出　　处：《计算机应用》2005年第9期2140-2142,共3页journal of Computer Applications

基　　金：国家自然科学基金资助项目(10476006)

摘　　要：Donoho等提出了一种基于小波变换的阈值除噪方法,这种方法在信号除噪及数据压缩等领域得到了广泛的应用。但是,Donoho设计的阈值处理方法对含有微弱的局部信号的除噪效果不明显。根据局部信号和白噪声有着不同的Lipschitz指数,经过小波变换后二者在时间-尺度平面上有不同特征,文中提出了一种新的除噪方法,利用该方法可以有效地去除噪声,同时保留微弱的局部信号,减少除噪过程中信号能量的损失。A method for de-noising by threshold, and their method has been used in many signal de-noising and compression problems. But it is not successful for the signals with detail local signal. It is known that detail local signals and noises have different Lipschitz coefficients, and after DWT they have different characters in time-scale domain. In this paper, a new improvement threshold method was proposed, which could de-noise effective and remain detail local, signals as many as possible.

关 键 词：小波变换 阈值处理 除噪 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]