基于多项式拟合的电容式传感器手势识别系统

Capacitive Sensor Gesture Recognition System Based on Polynomial Fitting

手势识别在人机交互方面具有重要的应用价值,但是传统设计方案中对用户的手势信息采集繁琐且设备昂贵,因此设计了一款基于多项式拟合的电容式传感器手势识别系统。系统采用FDC2214传感器读取数据代入到拟合好的多项式函数中,进而判别不同的手势,并利用无线通信模块将判别结果发送给PC监测终端。介绍了该系统的软硬件及算法设计,并给出了实验测试结果,结果表明:结合该算法的手势识别系统手势学习总时间为1.09 s,每个手势识别时间为0.62 s,准确率为97.41%,具有灵敏度高、准确性好,抗干扰性强的优点。

Gesture recognition has important application value in human-computer interaction.However,in traditional design solutions,the user’s gesture information collection is cumbersome and the equipment is expensive.Therefore,a capacitive sensor gesture recognition system based on polynomial fitting is designed.The system uses the FDC2214 sensor to read the data and substitutes it into the fitted polynomial curve function,in order to distinguish different gestures,and then sends the discrimination results to the remote PC monitoring terminal with the wireless communication module.The detailed designs about the software,hardware and algorithm are provided in this treatise.The experimental results show that the gesture learning time of the gesture recognition system combined with this algorithm is 1.09 s,the recognition time is 0.62 s,and the accuracy rate is 97.41%,which has the advantages of high sensitivity,good accuracy and strong anti-interference.