| 采用多特征融合的镜头边界检测方法 |
| |
| 引用本文： | 刘群,江伟,吴渝.采用多特征融合的镜头边界检测方法[J].计算机工程与应用,2010,46(13):171-174. |
| |
| 作者姓名： | 刘群 江伟 吴渝 |
| |
| 作者单位： | 重庆邮电大学计算机科学与技术学院,重庆,400065 |
| |
| 基金项目： | 重庆市自然科学基金No.2007BB2386;;重庆市教委应用基础研究项目(No.KJ070504)~~ |
| |
| 摘 要： | 提出了一种特征融合的镜头边界检测方法。HSV色彩直方图进行镜头边界检测是一种常用、有效的方法，该文融合了视频帧的HSV直方图特征、边缘特征和纹理特征，计算出不连续帧。最后，采用Kohonen网络自组织网络对不连续帧值进行聚类得到镜头边界。实验结果表明该方法不仅是可行和有效的，也解决了需要阈值的问题。
|
| 关 键 词： | 特征融合 镜头边界检测 自组织网络 色彩直方图 |
| 收稿时间： | 2008-10-21 |
| 修稿时间： | 2009-1-13
|
Approach of shot-boundary detection based on multi-feature fusion |
| |
| LIU Qun,JIANG Wei,WU Yu.Approach of shot-boundary detection based on multi-feature fusion[J].Computer Engineering and Applications,2010,46(13):171-174. |
| |
| Authors: | LIU Qun JIANG Wei WU Yu |
| |
| Affiliation: | School of Computer Science and Technology,Chongqing University of Posts and Telecommunications,Chongqing 400065,China |
| |
| Abstract: | An approach of shot-boundary detection by fusing feature is proposed.HSV histogram is a common and effective method for shot boundary detection.In this paper,discontinuous frame is obtained based on HSV histogram,edge orientation histograms and texture feature firstly.Then the Kohonen self-organized network is used to receive the shot boundary by clustering the discontinuous frame.The results of experiments show that this proposed approach can detect shot effectively and doesn't need threshold. |
| |
| Keywords: | feature fusion shot boundary detection self-organized network HSV histogram |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《计算机工程与应用》浏览原始摘要信息 |
|
点击此处可从《计算机工程与应用》下载全文 |
