| 一种新的基于对象的足球视频镜头分类方案 |
| |
| 引用本文： | 周艺华,曹元大,张龙飞. 一种新的基于对象的足球视频镜头分类方案[J]. 计算机工程与应用, 2005, 41(34): 229-232 |
| |
| 作者姓名： | 周艺华 曹元大 张龙飞 |
| |
| 作者单位： | 北京理工大学信息科学技术学院计算机科学工程系,北京,100081 |
| |
| 摘 要： | 论文提出了一种基于对象的足球视频镜头分类方案。首先对足球视频中的场地和运动员对象进行检测和分割,然后利用识别出的场地特征、运动员数目及运动员与场地比例等特征,对足球视频中的长距镜头、中距镜头、特写及其它类型的镜头进行分类。实验表明,该分类方案取得了良好的效果。
|
| 关 键 词： | 镜头分类 特征提取 视频摘要 |
| 文章编号： | 1002-8331-（2005）34-0229-04 |
| 收稿时间： | 2005-02-01 |
| 修稿时间： | 2005-02-01 |
A Novel Shot Type Classification Based on Object in Soccer Video |
| |
| Zhou Yihua,Cao Yuanda,Zhang Longfei. A Novel Shot Type Classification Based on Object in Soccer Video[J]. Computer Engineering and Applications, 2005, 41(34): 229-232 |
| |
| Authors: | Zhou Yihua Cao Yuanda Zhang Longfei |
| |
| Affiliation: | Department of Computer Science,School of Information Science of Technology, Beijing Institute of Technology,Beijing 100081 |
| |
| Abstract: | This paper presents a novel shot type classification scheme based on object in soccer video.Firstly,Field and Athlete objects are detected and segmented.And then long shot,medium shot,close-up shot and other shots are classified by the features such as field area,number of athletes and the ratio between athlete and field.Experiment shows that the scheme performs good performance. |
| |
| Keywords: | shot classification feature extraction video abstraction |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
