A Current-Based Surface Electromyography (sEMG) System for Human Motion Recognition: Preliminary Study

• Published in: Intelligent Robotics and Applications Vol. 13013; pp. 737 - 747
• Main Authors: Zeng, Cheng, Zheng, Enhao, Wang, Qining, Qiao, Hong
• Format: Book Chapter
• Language: English
• Published: Switzerland Springer International Publishing AG 2021; Springer International Publishing
• Series: Lecture Notes in Computer Science
• Subjects: Current-based sEMG; Human upper limb; Motion recognition; Stability
• ISBN: 9783030890940; 3030890945
• ISSN: 0302-9743; 1611-3349
• DOI: 10.1007/978-3-030-89095-7_70

The myoelectric interface acts as an important role in the field of wearable robotics. This study explores the current-based sEMG technology for upper-limb motion recognition. Different from the voltage-based sEMG system, the current-based sampling approach can directly extract the current signals and be free from the cross talks. The technology facilitate the myoelectric sampling in a non-ideal environment such as underwater. We designed the sensing circuit with a feedback-loop current amplification module, and analyzed the stability. After development of the system, three healthy subjects participated in the experiment. Six basic wrist joint motions were investigated. With the selected feature set and the designed classification method, The average recognition accuracies across the subjects were 96.3%, 94.2%, and 95.8% respectively. The preliminary results demonstrate that the current-based sEMG technology is a promising solution to upper-limb motion recognition in a rigorous environment.