Kinematics Parameter Analysis of Long Jump Based on Human Posture Vision

In order to enable long jumpers or coaches to accurately analyze each training result and help athletes improve their performance, this paper designs a method of analyzing the kinematic parameters of long jump based on human posture vision. Firstly, the attitude recognition model is used to analyze the athlete's posture, and the key point position information of the athlete is given. At the same time, the trajectory of the movement is accurately tracked through perspective transformation and background difference, and the kinematic parameters of the long jump are obtained based on the trajectory.