AR与VR技术在冬季体育赛事中的应用分析  

The application analysis of augmented reality and virtual reality technology in winter sports events

作　　者：牛鹤璇 王艾莎[2] 胡萍[1] 刘梅[2] NIU Hexuan;WANG Aisha;HU Ping;LIU Mei(Department of Leisure and Social Sports,Harbin Institute of Physical Education,Harbin 150040,China;Department of Sports Health and Humanities,Harbin Institute of Physical Education,Harbin 150040,China)

机构地区：[1]哈尔滨体育学院休闲与社会体育系,黑龙江哈尔滨150040 [2]哈尔滨体育学院体育健康与人文系,黑龙江哈尔滨150040

出　　处：《西安邮电大学学报》2019年第5期105-110,共6页Journal of Xi’an University of Posts and Telecommunications

基　　金：黑龙江省哲学社会科学研究规划资助项目(18TYD357);黑龙江省哲学社会科学研究规划资助项目(17TYB057);中国大学生体育协会“十三五”规划2017年度重点课题资助项目(201713505);哈尔滨体育学院校级课题资助项目(2017TJ011)

摘　　要：为了更加清晰和流畅转播冬季体育赛场上的实时画面,设计了一种智能视频分析系统。该系统应用迭代最近点(iterative closest point,ICP)算法对冬季赛场内运动员图像通过点云配准进行实时拼接,并且能够计算出单应性矩阵对原始场景进行增强,增强观众对增强现实(augmented reality,AR)技术与虚拟现实(virtual reality,VR)技术的交互式体验感受。利用平昌冬奥会美国队与韩国队冰球比赛小组赛录像视频,对本智能视频分析系统进行测试,结果表明,本系统可以准确跟踪和识别冬季赛场上运动员,提高了冬季赛场上运动员目标检测的精度。In order to rebroadcast winter sports competition more clearly and smoothly,an intelligent video analysis system is designed,which can use iterative closest point algorithm to joint real-time athlete images through point cloud matching and calculate the homography matrix to enhance the original scene,and increase AR and VR interactive experience.We tested this algorithm in the intelligent video analysis system by using the video of the pyeongchang winter Olympic Games group between the United States team and the Korea team.The test results showed that this method could track and identify athletes on the winter field,and improve the identification accuracy.

关 键 词：体育赛事 增强现实 虚拟现实 自动跟踪 目标识别 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术] G84[自动化与计算机技术—计算机科学与技术]