基于一种改进阈值函数的小波去噪方法研究  

Research on wavelet denoising method based on an improved threshold function

作　　者：张金武[1] 冯毅[1] 李文[1] 

机构地区：[1]华南理工大学化工机械与安全工程研究所,广东广州510640

出　　处：《电子设计工程》2017年第9期137-140,144,共5页Electronic Design Engineering

摘　　要：针对传统的阈值函数方法和阈值选取方法存在的一些不足,在现有的研究成果基础上,本文采用了一种改进的阈值函数方法,既克服了硬阈值函数的缺点,又减小了软阈值函数的偏差。通过MATLAB仿真,得出用本文方法去噪后的信噪比均在37.326以上,高于硬阈值函数的37.164和软阈值函数的37.265;均方差均在5.680以下,低于硬阈值函数的5.787和软阈值函数的5.720,说明本文改进的阈值函数方法去噪效果优于软、硬阈值函数,适用于含噪信号的分析和处理。In view of the disadvantages of the traditional threshold function method and the threshold selection method,this paper proposes an improved threshold function method based onexisting research results, which not only overcomes the shortcomings of the hard threshold function, but also reduces the deviation of the soft threshold function. By conducting MATLAB simulation experiment,we conclude that the SNR of the method used in this paper is more than 37.326, which is higher than 37.164 ofthe hard threshold function and 37.265 of the soft threshold function;the MSE is below 5.680, which is less than 5.787 of the hard threshold function and 5.720 of the soft threshold function,the result shows that denoising effect of the improved threshold function is better than the traditional methods,and it is suitable for the noisy signal analysis and processing.

关 键 词：小波阈值去噪 MATLAB仿真 信噪比 均方误差 

分 类 号：TN912.3[电子电信—通信与信息系统]