摘要
为了消除肌电信号中的噪声,提出了一种基于相邻尺度积系数的硬阈值滤波方法.通过对采集的肌电信号进行小波分解,并对肌电信号的各层噪声方差进行估计,构造一种基于相邻尺度积系数的硬阈值函数,实现了真实信号与噪声的分离.根据保留下的小波系数进行重构,得到滤波后的信号.实验表明该方法能有效消除噪声,且基本保留了真实信号的边缘特征,为提高基于肌电信号的手部动作识别率提供了技术手段.
To eliminate the noise included in electromyography(EMG),a de-noising method by multi-scale product coefficient hard thresholding was presented.The acquired EMG was decomposed using wavelet transform,and the noise square error of each layer was estimated.Via constructing a hard thresholding function by multi-scale product coefficient,the true signal and noise was separated.Finally,through the reconstruction of the reserved wavelet coefficients,the de-noised signal was obtained.The experimental results show ...
出处
《华中科技大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2008年第S1期102-104,共3页
Journal of Huazhong University of Science and Technology(Natural Science Edition)
基金
国家高技术研究发展计划资助项目(2008AA04Z212)
浙江省科技计划资助项目(2007C23088)
关键词
肌电信号
小波变换
噪声
阈值
相邻尺度积系数
electromyography（EMG）
wavelet transform
noise
thresholding
multi-scale product coefficient
