期刊
INTERNATIONAL JOURNAL OF HUMANOID ROBOTICS
卷 12, 期 2, 页码 -出版社
WORLD SCIENTIFIC PUBL CO PTE LTD
DOI: 10.1142/S0219843615500115
关键词
Multichannel sEMG; hand motion recognition; feature; differential root mean square; EMG electrode configuration
类别
资金
- UK Engineering and Physical Science Research Council [EP/G041377/1]
- National Basic Research Program (973 Program) of China [2011CB013305]
- Science and Technology Commission of Shanghai Municipality [13430721600]
- Engineering and Physical Sciences Research Council [EP/G041377/1] Funding Source: researchfish
- EPSRC [EP/G041377/1] Funding Source: UKRI
Surface electromyography (sEMG)-based hand motion recognition has a variety of promising applications. While a person performs different hand motions, commands can be extracted to control external devices, such as prosthetic hands, tablets and so forth. The acquisition of discriminative sEMG signals determines the accuracy of intended control commands extraction. This paper develops an 16-channel sEMG signal acquisition system with a novel electrode configuration that is specially designed to collect sEMG on the forearm. Besides, to establish the relationship between multichannel sEMG signals and hand motions, a 2D EMG map is designed. Inspired from the electromyographic (EMG) map, this paper proposes an EMG feature named differential root mean square (DRMS) that somewhat takes the relationship between neighboring EMG channels into account. In the task of four hand motion discrimination by K-means and fuzzy C-means, DRMS outperforms traditional root mean square (RMS) by 29.0% and 36.8%, respectively. The findings of this paper support and guide the use of sEMG techniques to investigate sEMG-based hand motion recognition.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据