Title: Classification of hand postures using forearm perimeter sensor and compensation of residual muscle volume change with sEMG
Authors: Hwiyong Choi; Sangyoon Lee
Addresses: Department of Mechanical Design and Production Engineering, Konkuk University, 120 Neungdong-ro, Gwangjin-gu, Seoul, Korea ' Department of Mechanical Design and Production Engineering, Konkuk University, 120 Neungdong-ro, Gwangjin-gu, Seoul, Korea
Abstract: This paper reports that performance of a sensor for measuring perimeter change of the forearm has been improved by compensating for residual muscle volume change. A key idea of the compensation is removing offset signal during the muscle relaxed state. One channel of surface electromyography (sEMG) signal was employed for detecting the muscle activity. The suggested method was validated by online classification of three different hand postures with one forearm perimeter sensor. The result shows that the suggested method compensates for residual muscle volume change effectively and improves accuracy of hand posture classifier.
Keywords: hand posture classification; residual muscle volume change; RMVC; forearm perimeter change; surface electromyography; sEMG; k-nearest neighbour; KNN classifier; human-computer interaction; HCI; sensor performance; hand postures; muscle activity.
International Journal of Mechatronics and Automation, 2014 Vol.4 No.4, pp.213 - 221
Received: 05 Nov 2013
Accepted: 14 Mar 2014
Published online: 17 Dec 2014 *