A Gesture Recognition System Based On MEMS Accelerometer

With the development of mobile intelligent electronic devices, gesture recognition based on MEMS accelerometers has become a research hotspot. This paper proposes a gesture recognition system that utilizes an ARM processor, ADXL345 accelerometer, NRF24L01 wireless communications module, and LABVIEW. The system incorporates five accelerometers located on fingers to collect acceleration information from each sensor. The collected acceleration data is filtered and compared with pre-defined thresholds to determine the state of each finger. The system wirelessly transmits the status information of the five fingers to a receiver unit through the wireless communications module. The receiver unit analyzes the finger states and generates a final gesture identification value, which is then transmitted to the LABVIEW upper computer via a serial port. The generated gesture identification value is matched with image templates in LABVIEW to achieve gesture recognition. Additionally, the system enables human-computer interaction by allowing users to control web page access, image display, and robotic gripper movement through gestures.