田径运动终点裁判的智能视觉分析研究

发布时间：2018-10-21 10:21

【摘要】：在裁判田径比赛的名次的过程中,由于运动速度过快,视觉对运动员的终点冲刺进行分析时,容易产生对终点冲刺中的像素拖拉延迟问题,如采用高频采集的方式进行解决,以最小采样周期消除干扰,过小的采样频率无法建立帧差特征,导致根据帧差特征的判断方法失效。提出一种改进冗余小波变换域算法的田径运动终点裁判智能视觉判断方法。对运动员冲刺过程视频图像进行预处理,将各帧图像中运动员冲刺终点过程与背景有效的分离,构造运动员冲刺终点特征的不变量,利用相似函数计算相邻帧冲刺终点特征相似度。利用冗余小波变换在小波域提取终点冲刺的区域,采用mean-shif算法利用运动员冲刺终点的色彩分布信息描述其终点冲刺的变化过程。仿真结果证明,上述方法能对比赛名次作出准确的裁判。
[Abstract]:In the process of judging the ranking of track and field competition, because of the speed of movement is too fast, when the visual analysis of the finish sprint of athletes, it is easy to produce the problem of pixel dragging delay in the final sprint, such as using the method of high frequency acquisition to solve the problem. When the interference is eliminated by the minimum sampling period, the frame difference feature can not be established when the sampling frequency is too small, which results in the failure of the judgment method based on the frame difference feature. This paper presents an intelligent visual judgment method for track and field final judgment based on improved redundant wavelet transform domain algorithm. The video image of sprint process is preprocessed, and the process of sprint finish point is separated from the background effectively in each frame image, and the invariant of athlete sprint finish point feature is constructed. Similarity function is used to calculate the similarity of the sprint endpoint features of adjacent frames. Redundant wavelet transform is used to extract the region of final sprint in wavelet domain, and mean-shif algorithm is used to describe the changing process of finish sprint by using the color distribution information of athlete's sprint finish. The simulation results show that the above method can make accurate referees.
【作者单位】：山东理工大学体育学院;
【分类号】：TP391.41;G82

【参考文献】