基于超声波的手势识别设备的研究

摘要

设计了一套基于主动超声技术的低成本、低功耗的手势识别装置,可以用来实现无接触式的人机交互.采用STM32为主控核心,发射超声波并使用4通道超声性能稳定的麦克风接收经过物体反射后的回波信号.在单片机内使用改进的过零检测法估算回波频率,并将回波的频率和幅值作为手势识别的特征,使用USB或串口传输至上位机.经验证,特征信息可反映用户手在垂直方向和水平方向的运动.%A low-cost and low-power gesture recognition device based on active ultrasound technology is designed to achieve non-contact human-computer interaction.The device uses STM32 as the microcontroller to transmit ultrasound signals to the target and receive the response by 4-channel ultra-stable MEMS microphone. A modified zero-crossing detection method is used to measure frequency on the microcontroller,and the frequency and amplitude of response signal are transmitted as the feature of gesture to upper computer through USB or serial port.The experimental results show that the feature can reflect the movement of target.