基于表面肌电控制的虚拟人机交互系统

Human-machine Interaction System Based on Surface EMG Signals

实现了一种基于肌电信号手势识别的实时虚拟人机交互系统。系统采集受试者执行六类不同手势动作时的上肢肌肉表面肌电信号,对其进行模式识别,并将识别结果作为虚拟物体的控制信号,控制虚拟飞机执行三个自由度的飞行动作,从而实现人与虚拟环境的友好交互。对10位受试者进行的单用户和多用户控制测试实验表明,实现的虚拟人机交互系统可实现基于肌电信号的虚拟人机交互,且具有虚拟环境生动形象、控制准确率高等特点。

A real-time virtual human-machine interaction system was proposed based on the surface EMG-based gesture recognition technology. Surface EMG signals were collected from the upper limbs of the subject when they did six defined hand gestures. Then surface EMG pattern recognition algorithm was conducted on them and the results were used as input commands to control 3-DOF （degree of freedom） movements of the virtual plane in a virtual environment. Single-user and multi-user online test experiments were conducted on 10 subjects to examine the system＇s reliability. The experiment results show that this system performs well in the surface EMG-based gesture control of the virtual plane with high accuracy.