Research On Motion Recognition Algorithm And Software Development For Student Group

In recent years, students’ fitness and health status continually decline. Low and high intensity exercise will have a bad effect on student’ physical fitness and health. How to effectively monitor and obtain quantitative physical activity status of students have been a serious problem, which need be solved by education administrators urgently. There are some problems in motion recognition products, such as complex operation, connection with smart phones. Their motion recognition function is not suitable for student group. In view of the above questions, this thesis mainly carries out a research on motion recognition algorithm, and designs and implements some software to monitor the exercise intensity of students. The main work and contribution of this thesis are as follows:Firstly, the significance of my research work is proposed. The present research status of human motion detection and the disadvantages of wearable products on the market are analyzed.Secondly, to deal with the problem that it is hard to distinguish walking, going upstairs and going downstairs, a recognition algorithm based on features such as variance, quartile deviation of the X-axis and X-axis skewness is proposed. The experimental results indicate that the extracted features can effectively distinguish walking, going upstairs and going downstairs. This thesis also proposes and implements an exercise intensity classification algorithm to evaluate the exercise intensity of human body in the exercise.Thirdly, a wearable device supplied by button battery is designed and implemented. The exercise intensity classification algorithm is implemented in the wearable device. By utilizing the BLE stack from TI supports, a wearable device software is implemented for calculating and saving the exercise intensity, then transmitting them to the transfer station via Bluetooth 4.0. The results of power test show that power consumption for a day is 0.4787 mAh.Fourthly, a transfer station is designed and implemented. By utilizing the BLE stack from TI supports and STM32 platform, a transfer station software is implemented for receiving, storage and retransmission of activity information. The experimental results show the transfer station runs reliably.The final chapter summarizes this thesis and proposes some prospects of the future research and development.