Behavior tracking and recognition system of basketball players based on machine vision
In the tracking technology employed to monitor the athletes'behavioral trajectory in basketball games,there are problems such as significant recognition error and low efficiency.In light of this,a more effec-tive method for tracking and recognising the basketball players'behavior based on machine vision technology is proposed.A basketball player behaviour acquisition sub-system based on machine vision is designed to implement accurate pre-processing of frame images captured by the camera for player behaviour profile recog-nition;and a Gaussian mixture model is used to initially extract basketball players'contour information,detect behavioral edge contour corner points to complete the optimization of edge contour feature extraction.The con-tour point distribution histogram is further obtained,based on which the pyramid matching algorithm is used to identify the behavioral trajectory of the athlete.The test results show that the false detection rate of this method for tracking athletes'behavior is approximately between 0.4％and 1.3％,and the leakage curve exhibitss a horizontal tendency,with a maximum value of less than 1％,which enhances the practical implimentaion of tracking and identifying the basketball players'behavior.
万方数据
维普
