Robust myoelectric interface for postural change, sweating and muscle fatigue based on spatio-temporal warping of time-series EMG signals

2014 Fiscal Year Final Research Report

• Project/Area Number: 25540085
• Research Category: Grant-in-Aid for Challenging Exploratory Research
• Allocation Type: Multi-year Fund
• Research Field: Human interface and interaction
• Research Institution: Nara Institute of Science and Technology
• Principal Investigator: MATSUBARA Takamitsu 奈良先端科学技術大学院大学, 情報科学研究科, 助教 (20508056)
• Project Period (FY): 2013-04-01 – 2015-03-31
• Keywords: 筋電ロボットインタフェース / 表面筋電位 / 動作識別 / 姿勢変化 / 時系列カーネル

Outline of Final Research Achievements

In this study, we consider the development of myoelectric human-robot interfaces that enable intuitive control of complex robots based on the user intention estimated from myoelectric signals. More concretely, we propose a motion intention recognition method based on the surface Electromyography (SEMG), robust to the factors of 1) postural change, 2) sweating and 3) muscle fatigue, all of which can largely change the relationship between the sEMG signal and the motion intention. Our method is achieved with using pattern matching techniques, and without either other sensors for measuring the postural change, the sweating and muscle fatigue, or physiological models of human body. Through the subjective experiment under the postural change, the proposed method resulted in better performance than that of previous methods.

Free Research Field

知能ロボティクス