一种新阈值函数的小波去噪算法研究  

Wavelet Denoising Algorithm based on a New Threshold Function

作　　者：陈映竹[1] 王玉文[1] 杨巍[1] 刘影[1] 

机构地区：[1]电子科技大学航空航天学院,四川成都611731

出　　处：《通信技术》2017年第7期1407-1411,共5页Communications Technology

摘　　要：小波变换由于能将信号时频特征局部化而被广泛应用于信号去噪。介绍小波阈值去噪中传统的硬阈值和软阈值去噪方法,提出一种新阈值函数小波去噪方法。该方法能有效弥补软、硬阈值去噪方法的不足,且克服了采用硬阈值法去噪导致信号振荡和软阈值法引入固定偏差的缺点。通过仿真验证表明,新的阈值函数有效提高了去噪效果,且在均方根误差、信噪比和平滑度上明显优于传统小波阈值去噪方法。Wavelet transform is widely used in signal denoising for its localizing the time-frequency characteristics of the signal. The traditional hard threshold and soft threshold denoising methods in wavelet threshold denoising are described, and new threshold function wavelet denoising method is proposed. This method can effectively compensate for the weakness of the soft and hard threshold denoising methods, and overcome the shortcoming in using the hard threshold method to reduce the signal oscillation and introducing the fixed deviation by the soft threshold method. Simulation results indicate that the new threshold function can effectively improve the denoising effect, and is superior to the traditional wavelet threshold denoising method in root mean square error, signal-to-noise ratio and smoothness.

关 键 词：小波变换 小波阈值去噪 阈值函数 均方根误差 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]