摘要
表面肌电信号是非平稳的随机信号,通过Mallet算法对其进行多分辨率分解分析,发现信号和噪声的小波变换系数具有相反的特性。研究了一种改进的自适应小波阈值算法,引入一个变量参数,将传统的统一阈值函数变为一种新的分层阈值函数;使阈值选择函数能够在闭区间上最大程度地保留有用信息。实验结果表明,该方法兼顾了软、硬阈值算法的优点,在表面肌电信号的视觉效果、信噪比和均方误差方面都有较好的效果。
Surface electromyography signals are non-stationary random signals. Multi-resolu-tion decomposition analyses of Mallet signals show that wavelet coefficients of signals and noises have opposite characteristics. This paper proposes an adaptive wavelet threshold algorithm using a threshold selection method with a new variable. The threshold function retains useful information in a closed interval to a maximum extent. The proposed method takes into account the advantages of both soft and hard threshold algorithms. Experimental results show that it has good visual effects, signal-t -noise ratio and mean square error.
出处
《上海电机学院学报》
2017年第4期215-219,共5页
Journal of Shanghai Dianji University
关键词
小波变换
表面肌电信号
阈值
wavelet transform
surface electromyography signal（SEMG）
threshold
