首页 | 本学科首页   官方微博 | 高级检索  
     

基于仿射传播聚类的视频关键帧提取方法
引用本文:谢晓能,吴飞.基于仿射传播聚类的视频关键帧提取方法[J].计算机应用与软件,2009,26(5).
作者姓名:谢晓能  吴飞
作者单位:1. 杭州广播电视大学,浙江,杭州,310012
2. 浙江大学计算机科学与技术学院,浙江,杭州,310027
基金项目:国家自然科学基金,浙江省自然科学基金 
摘    要:关键帧提取是基于内容的视频摘要生成中的一个重要技术.首次引入仿射传播聚类方法来提取视频关键帧.该方法结合两个连续图像帧的颜色直方图交,通过消息传递,实现数据点的自动聚类.并与k means和SVC(support vector clustering)算法的关键帧提取方法进行了比较.实验结果表明,AP(Affinity Propagation)聚类的关键帧提取速度快,准确性高,生成的视频摘要具有良好的压缩率和内容涵盖率.

关 键 词:视频摘要  关键帧提取  仿射传播聚类

VIDEO KEY FRAME EXTRACTION USING AFFINITY PROPAGATION CLUSTERING
Xie Xiaoneng,Wu Fei.VIDEO KEY FRAME EXTRACTION USING AFFINITY PROPAGATION CLUSTERING[J].Computer Applications and Software,2009,26(5).
Authors:Xie Xiaoneng  Wu Fei
Affiliation:Hangzhou Radio and TV University;Hangzhou 310012;Zhejiang;China;College of Computer Science and Technology;Zhejiang University;Hangzhou 310027;China
Abstract:Key frame extraction has been recognized as an important technique in content-based video summarization generation.In this paper,affinity propagation clustering is introduced for the first time in key frame extraction,this new approach clusters data points automatically through messages transmission based on combining colour histogram crosses of two serial image frames.It has been compared with the key frame extraction methods of k-means and support vector clustering algorithm,experimental results show that...
Keywords:Video summarization Key frame extraction Affinity propagation clustering  
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号