本文已被:浏览 1140次 下载 689次
投稿时间:2013-03-20
投稿时间:2013-03-20
中文摘要: 给出了基于霍夫变换的平行线检测算法,能初步定位球门.对二值图像竖直方向的目标点进行累加,并运用直方图分析,对球门作进一步定位.利用三基色直方图进行镜头分割,并用基于平行线检测的球门定位算法进行精彩片断的选择.经过对多个视频进行精确定位的结果表明,球门检出率可达90%以上.
Abstract:The detectionof parallel lines based on the Hough transform is proposed to calculate the initial positioning for the goal.Furthermore,the value of every pixel on the whole image is accumulated along the vertical direction to get the accurate location of thegoal.Finally,three-color histogram is designed for shot segmentation,and the proposed frame work is used for selecting RGB color highlights.The general detection of soccer goal video experimental results show that therate of detection goal is higher than90%
文章编号:20130324 中图分类号: 文献标志码:
基金项目:
引用文本:
仝明磊,李毅.足球比赛视频球门检测[J].上海电力大学学报,2013,29(3):303-305.
TONG Minglei,LI Yi.Goal Detection In Soccer Video[J].Journal of Shanghai University of Electric Power,2013,29(3):303-305.
仝明磊,李毅.足球比赛视频球门检测[J].上海电力大学学报,2013,29(3):303-305.
TONG Minglei,LI Yi.Goal Detection In Soccer Video[J].Journal of Shanghai University of Electric Power,2013,29(3):303-305.