基于表面肌电信号的虚拟现实控制系统设计

Design of control system for virtual realitybased on surface electromyography signal

摘要:

为了减少传统康复训练的单调乏味,设计了一种利用无线表面肌电控制虚拟现实环境中厨房的实时控制系统。该系统是以MATLAB为平台,利用图形用户界面进行设计；在算法上利用均方值与移动平均窗方法对实时肌电信号进行检测与分割；采用幅值绝对值均值和小波系数奇异值作为特征值；选取支持向量机算法进行模型训练和分类识别；最后,完成对虚拟厨房的动作控制。结果表明,该系统可以完成对虚拟厨房动作的实时控制,平均识别率为90.31%。后续可用于肌肉康复训练的患者,可提供具有沉浸感和真实感的虚拟厨房生活场景,对患者的康复具有积极意义。

Abstract:

In order to reduce the tedium of traditional rehabilitation training, this paper designed a real-time control system using wireless surface electromyography signal to control kitchen in virtual reality environment. This system was built in graphical user interface which based on MATLAB. It used the mean square value and moving average window method detected and segmented real-time surface electromyography signal. It used the mean absolute value of the amplitude and singular values of wavelet coefficient as feature vectors. It used support vector machine to train model and identify action. Finally, it completed the control of the virtual kitchen. Experiment results show that the system can complete the control of virtual kitchen actions and the average recognition rata is 90.31%. This system can be used for patients who need muscle rehabilitation training, and provide them with an immersive and realistic virtual kitchen living scenes. So it has positive significance for the rehabilitation of patients.