| Summary: | In order to make up for the lack of existing products on the market for identifying table tennis and badminton sports state, this paper designs a racket wrist which can identify these two kinds of movement in a variety of state watches. It′s hardware mainly uses STM32F103C8T6 single chip, MPU6050 sensor, and Bluetooth serial module. The method of motion state recognition is as follows：firstly，the three-axis acceleration and the triaxial angular velocity are used to calculate the triaxial angle. The data are subjected to high-pass filtering, smoothing, and data preprocessing. And then according to the random forest classification recognition algorithm, a variety of states of table tennis and badminton in the process are identified. The watch has a low cost of hardware, small size, and low power consumption, and by test the recognition rate is up to 90%.
|
|---|