Automobile Motion Posture Monitoring Based on Stereo Vision Technology

摘要

In the athletes' automobile motion process, degree of human joint movement is relatively complex, and the movement speed is very fast. The subtle changes between joints are very difficult to be constrained by fixed constraint algorithms. Traditional static image modeling method using a fixed constraint model is very hard to meet this rapid change of movements' key characteristics in dynamic small area, which makes it difficult to observe the changes of relative automobile motion parameters. This paper presents an athlete stereo vision calibration posture test method to collect automobile motion parameters through dual camera in athlete movement process. Complete the calibration of movement parameters through the perspective projection and camera parameters through effective coordinate change, calculate the changes of three dimensional automobile motion characteristic parameters, and complete the automobile motions' three dimensional parameters measurement by using factorization method. The simulation experimental results show that the method can measure athletes' three dimensional automobile motion parameters of high precision, and improve parameters' measuring precision.