文献类型：期刊文献

英文题名：Research on Chinese sign language recognition methods based on mechanomyogram signals analysis

作者：Feng, Wanjun[1]; Xia, Chunming[1]; Zhang, Yue[1]; Yu, Jing[1]; Jiang, Wendu[1]

机构：[1] School of Mechanical and Power Engineering, East China University of Science and Technology, Shanghai, China

年份：2019

起止页码：46

外文期刊名：2019 IEEE 4th International Conference on Signal and Image Processing, ICSIP 2019

收录：EI(收录号：20194507628553)

语种：英文

外文关键词：Wavelet analysis

摘要：This paper presents an integrated approach to Chinese sign language (CSL) actions recognition, which involves Teager-Kaiser energy operator (TKEO) segmentation, wavelet feature extraction and support vector machine (SVM) classification on mechanomyogram (MMG) Signals. It used a four-channel wireless signal acquisition system to collect the MMG signals of the extensor digitorum (ED), flexor carpi radialis (FCR), flexor carpi ulnaris (FCU) and extensor carpi radialis (ECR). After filtering, the TKEO algorithm was used to segment the MMG signals. The wavelet packet energy (WPE) of MMG signals were extracted as features for further analysis. SVM was applied as a classifier to recognize 18 CSL actions. Compared with other commonly used methods, the proposed method had better recognition accuracy and recognition performance as well. The average recognition accuracy of the proposed method was up to 95.38%. ? 2019 IEEE.