模拟篮球比赛中犯规镜头扫描自动检测技术研究  

Simulation of the Foul Action Lens Scans Automatic Detection in the Basketball Match

作　　者：谢天虹 XIE Tian-hong(Henan Technical College Of Construction,Zhengzhou 450007,China)

机构地区：[1]河南建筑职业技术学院,河南郑州450000

出　　处：《广州体育学院学报》2020年第5期106-112,共7页Journal of Guangzhou Sport University

摘　　要：利用犯规动作检测技术准确捕捉和检测篮球比赛中的犯规动作,是在对抗激烈的比赛中精准判罚的重要辅助技术。通过实验研究,提出一种犯规镜头扫描自动检测技术。该技术采用FGRG算法获取篮球运动员的动作轮廓,并利用AdaBoost算法提取篮球运动员动作图像的关键帧,最后利用LLE算法提取特征信息来判断动作是否犯规。实验证明,该技术有效地提高了激烈对抗的篮球比赛中运动员的犯规动作自动检测识别的准确度。To accurately detect fouls in the basketball game,it is necessary to study the technique of foul action lens scan automatic detection in the basketball match.Using the current technique of foul motion detection in the basketball game is difficult to detect foul action of basketball players in a heated game.For this reason,a foul action lens scan automatic detection technology in a basketball game is raised in this paper.This technique uses the FGRG algorithm to extract the basketball player's action contour and uses the AdaBoost algorithm to extract the keyframe of the basketball player's action image.Finally,the characteristic information is extracted by LLE algorithm to determine whether the action is foul.The experiment proves that the technique can accurately detect the foul action of basketball players in the game.

关 键 词：篮球比赛 犯规动作 扫描自动检测 

分 类 号：G814[文化科学—体育学]