华中科技大学学报（自然科学版）2011，Vol.39Issue(z2)：153-156,4.
基于表面肌电信号的时频组合特征融合识别
A recognition method of multi-channel SEMG based on the fusion of time-frequency combination characteristics
摘要
Abstract
In order to overcome the onesidedness of single feature describing in single domain, a method of time-frequency combination feature extraction and fusion was proposed. Selecting the signal' s three characteristics,skewness and kurtosis in time domain, the ratio of power spectrum in frequency domain and the wavelet coefficients in time-frequency domain,then analyzing and processing each characteristics intensively. Through the BP network multi-step classification, using fuzzy integration to process the result so as to accumulate the evidence for the final judgment. The experimental results indicate that through the fusion of multi-characteristics, the electromyography's feature gets more comprehensive expression, the recognition rate of hand motions that based on the multi-channel semg increased remarkably.关键词
表面肌电信号/多特征融合/时频组合/BP网络/模糊积分Key words
surface electromyography (SEMG)/ multi-characteristics fusion/ time-frequency combination/ BP network/ fuzzy integration分类
信息技术与安全科学引用本文复制引用
杨新亮,罗志增..基于表面肌电信号的时频组合特征融合识别[J].华中科技大学学报（自然科学版）,2011,39(z2):153-156,4.基金项目
国家自然科学基金资助项目(60874102). （60874102）
