Research and realization of crouch start correction system based on human pose estimation

For beginners in sprinting, receiving timely and effective guidance and feedback can not only help them master the movement, but also effectively avoid sports injuries. However, the existing guidance and improvement methods are inconvenient and inaccurate. This paper makes an in-depth study of the human body posture synthesis model based on deep neural network, and uses it to obtain posture and joint data of athletes, and judges whether the posture is correct and standard through the correction system. It is found that the model in this paper can obtain more accurate human body data. The position correction system has a higher posture recognition rate, which can effectively judge whether the position posture is standard or not, it is of great significance for actual position guidance.