Prediction method of running fatigue based on depth image

The number of people who participate in running every year is huge, and in the United States alone, at least 60 million people participate in running every year. At the same time, the problem of stress fractures caused by running is also very common. The biggest cause of this injury is over fatigue exercise. To prevent over-fatigue exercises that may occur during running and accurately capture every small posture change of runners, this paper proposes a running fatigue prediction method based on the depth image. The method uses the Borg scale (RPE) as the label for measuring fatigue and the spatial information of joint points as the data describing the running posture to make a prediction of the current fatigue state of the tester, which truly achieves the three practical requirements of contact-free, non-interference and timely processing feedback for running fatigue prediction.