关于长跑运动员起跑图像姿态优化校正仿真

Simulation and Optimization of Starting Posture of Long Distance Runners

对长跑运动员的起跑图像姿态的优化矫正,能够有效提高运动员跑步效率。对运动员起跑姿态图像的矫正,需要计算相同位置像素块的绝对差分,给出长跑运动员起跑姿态合理性匹配阈值,完成对起跑姿态优化矫正。传统方法先建立区域像素点之间的联系,生成起跑姿态图像的高斯模型,但忽略了给出姿态合理性匹配阈值,导致姿态矫正精度偏低。提出基于朗斯基函数的图像分析下长跑运动员起跑姿态优化校正方法。利用三维图像分析原理给出运动员起跑姿势的区域与背景显露区域,通过朗斯基矩阵行列式判断相邻像素间空间域相关性,计算当前帧中搜索窗口中心位置的待匹配像素块与参照帧中相同位置像素块的绝对差分,给出长跑运动员起跑姿态合理性匹配阈值,完成对长跑运动员起跑姿态优化校正。仿真证明,所提方法矫正效果较为明显,对运动员专项技术动作的形成与发挥有着积极意义。

An optimization correction method for starting posture of long distance runner is proposed based on Ion- ski function under image analysis. Firstly, the area and background exposure area of an athlete ＇ s starting position is given by utilizing principle of three-dimensional image analysis. Through lonski matrix determinant, spatial correla- tion between adjacent pixels is judged, and then pixel block to be matched of central location of search window in cur- rent frame and absolute difference of pixel block at the same location in reference frame is calculated. Finally, the ra- tional matching threshold of starting posture of long distance runner is obtained, and the optimization correction of starting posture of long distance runner is completed. Simulation results show that the proposed method has more obvi- ous correction effect. It plays a positive role on formation and exertion of athletes＇ special technical movement.