计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (24): 72-74.

• 学术探讨 • 上一篇    下一篇

基于新阈值函数及最优尺度的小波去噪研究

刘恒冰,韩世勤,刘 晶

  

  1. 中国地质大学 数理学院,武汉 430074
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-08-21 发布日期:2007-08-21
  • 通讯作者: 刘恒冰

Wavelet de-noising based on novel thresholding function and best decomposition scale

LIU Heng-bing,HAN Shi-qin,LIU Jing   

  1. School of Mathematics and Physics,China University of Geosciences,Wuhan 430074,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-08-21 Published:2007-08-21
  • Contact: LIU Heng-bing

摘要: 在Donoho D L和Johnstone I M提出的小波阈值去噪算法的基础上,首先构造了一种新的阈值函数。与传统软、硬阈值函数相比,新阈值函数不但连续,而且高阶可导,克服了硬阈值函数不连续及软阈值函数中小波估计系数与分解系数之间存在恒定偏差的缺陷。同时,为了获得更好的去噪效果,提出了基于白噪声?字2检验确定小波最优分解尺度的方法。最后,通过数值仿真实验,证明了基于白噪声?字2检验方法的有效性;在最优分解尺度下,新阈值函数在信噪比增益和最小均方误差意义上均优于传统阈值函数。

关键词: 小波变换, 阈值函数, 分解尺度, 均方误差, 信噪比

Abstract: A novel thresholding function is presented firstly based on the wavelet thresholding de-noising algorithm put forward by Donoho D L and Johnstone I M.Comparing with soft- and hard-thresholding function,the new thresholding function is not only continuous,but also has a high order derivative.It overcomes the shortcomings of conventional thresholding functions,such as discontinuous of hard-thresholding and the invariable dispersion in soft-thresholding.Second,a method to determine the best decomposition scale via white noise ?字2 verification is presented.At last,simulation results indicate that the above method is effective and the new thresholding function gives better MMSE performance and SNR gains than conventional thresholding function.

Key words: wavelet transform, thresholding function, decomposition scale, MSE, SNR