基于DSP和MEMS的人体动作识别系统

System of Human Motion Detection Based on DSP and MEMS

为解决现有虚拟训练软件的人体动作输入设备沉浸感和训练效果不足的问题,采用DSP和微机电系统(MEMS)构建人体动作识别系统。该系统应用MEMS器件为核心构建微惯性测量单元(MIMU)模块,模块的DSP芯片完成传感器数据采集和人体姿态解算。实验结果表明系统样机在动态活动测试中运行稳定可靠,动态解算精度在1°以内,可准确测量人体动作,满足虚拟训练要求。该研究提供了一种新型的虚拟训练软件的硬件输入设备,具有一定的市场价值。

To improve the sense of immersion and training effects of human detection input device for the virtual training software,the system of human motion detection based on DSP and MEMS was designed.The system uses MEMS as key components to form the Micro Inertial Measurement Unit（MIMU）,and the DSP chip of the MIMU samples the sensor＇s voltage values and makes the attitude algorithm.Experiments results show that the system runs stable in motion test and resolving precision can up to 1 degree at least.Thus the system can detect human motion accurately and meet the requirement of virtual software training.The research provides a new hardware design of input signal device for virtual training software and then has a certain market valuation.