基于加速度计的手指运动姿态检测

针对动态测量手指关节角度的需要,设计了一种基于MEMS加速度计的手指运动姿态检测方法。在该方法中,通过固定于手指上的三轴加速度计ADXL330的各个敏感轴感受到的重力加速度分量的大小来检测手指的关节角度。为了验证该方法的可行性,设计了一步进电机控制的能在竖直平面内转动的装置,通过同步检测传感器的输出和电机转动的角度,评估传感器在动态条件下的测量精度,结果表明:在动态条件下,其绝对误差为1°～2.5°,相对误差为3%～6%。在此基础上,将传感器用于测量手指关节的运动姿态,采集手指敲击键盘时传感器的输出,通过数据处理获得食指MCP关节、PIP关节和DIP关节随时间的变化关系。实验结果表明:加速度计可以有效地检测手指的运动姿态。

Detection of finger motion gesture based on accelerometer

To meet the need of dynamic angle measurement of finger joint,a method for finger motion gesture detecting is designed based on MEMS accelerometer.In this method,the joint angle is determined by the component of the gravity acceleration sensed by the sensitive axes of accelerometers.To verify the feasibility of the proposed scheme,a rotating device which can be controlled by the stepper motor is designed in vertical plane,and the measurement precision of the sensor in dynamic conditions is evaluated by detecting the sensor output and the motor rotation angle synchronously.The results show that the absolute error and relative error mainly range of 1°～2.5° and 3 %～6 % respectively in the dynamic testing.On the basis of this,the accelerometer is used to detect the motion gesture of finger joint when the finger is tapping keyboard,and acquire the angle variation with time of the index finger’s MCP joint,PIP joint and DIP joint by processing datas collected by the sensor.The preliminary results indicate that accelerometer-based finger motion detection is an effective method.