基于Meanshift算法的110 m跨栏运动员助跑轨迹跟踪方法

Run-up Track Tracking Method of 110-meter Hurdle Athletes Based on Meanshift Algorithm

为充分发挥助跑对提升110 m跨栏成绩的促进作用,提出了基于Meanshift算法的110 m跨栏运动员助跑轨迹跟踪方法。使用摄像机采集110 m跨栏运动员助跑视频图像,创建110 m跨栏运动员助跑轨迹跟踪的运动学模型,实现110 m跨栏运动员助跑轨迹跟踪。实验结果表明:该方法对各帧图像内运动员位置的预测均方根误差始终低于0.35×10^(-5);该方法均具有较理想的助跑轨迹跟踪能力,且运动员的起跨前最佳成绩可达到8.5m/s。

In order to give full play of promoting effect of run-up on improving the performance of 110 m hurdle,a method of track tracking of 110-meter hurdle athletes’run-up based on Meanshift algorithm is proposed.A camera is used to collect the video image of 110-meter hurdle athletes’run-up and a kinematic model is established to realize the track tracking.The experimental results show that the mean root square error of prediction of athletes’position in each frame is always lower than 0.35×10^(-5).This method has better effects for run-up track tracking,and the best performance of athletes before the starting of the stride can reach 8.5 m/s.