基于随机抽样一致性算法的运动平面上点的视觉测量及其应用

本文发表在《中国体育科技》2008年第1期

王岚,沈乐君,柯遵渝. 基于随机抽样一致性算法的运动平面上点的视觉测量及其应用. 中国体育科技,2008,1

使用计算机视觉技术，对球场中运动员和关键点的位置进行测量，是基于视频的技战术分析系统的核心技术，也是运动解析的重要原理，更是体育科学研究的重要工具。本文介绍了其理论基础与实现方法，探讨了测量误差和RANSAC容错算法，介绍了使用三维模型进行标定和结果演示。大量实验证明，本文提出的方法和传统方法相比，提高了纠错能力和适应性，在体育科研中有广泛的应用价值。

Abstract: computer vision based visual measurement of points on ground plane via video or digital image are introduced by paper. What we presented is fundamental for sport software and crucial for the tactic analysis. The paper introduces the basic theories; discuss the error of measurement and use RANSAC algorithm framework to improve algorithm accuracy. The 3D model can visible and very useful when we do calibration and presentation of analysis. The results of experiments showed that the method in the paper is more robust, flexible than traditional method and can be widely used in many fields of sport science.

演示图片

全文下载：

点击此处下载(文件大小422KB)

关键词：视觉测量;RANSAC;单应性矩阵;图像坐标;球场坐标;球落点;网球;运动员跟踪;

Keywords: visual measurement; RANSAC; homography matrix; image coordinates; stadium coordinates; serve placement; tennis; athlete tracking;