Computer Engineering and Applications ›› 2009, Vol. 45 ›› Issue (15): 9-10.DOI: 10.3778/j.issn.1002-8331.2009.15.003

• 博士论坛 • Previous Articles     Next Articles

Combining principal components analysis and clustering to extract key-frame

XU Wen-zhu,XU Li-hong   

  1. School of Electronics and Information,Tongji University,Shanghai 200092,China
  • Received:2009-02-03 Revised:2009-03-09 Online:2009-05-21 Published:2009-05-21
  • Contact: XU Wen-zhu

结合主成分分析和聚类的关键帧提取

许文竹,徐立鸿   

  1. 同济大学 电信学院,上海 200092
  • 通讯作者: 许文竹

Abstract: Key-frame extraction is important to content-based video analysis.For extracting key frames efficiently from different video,this paper presents an efficient method for key-frame extraction in which principal components analysis is applied to key-frame extraction.The initial clustering centers and the number of clustering k are determined by visual content,through k-means clustering get the key frames.The experiment indicates the algorithm can filter out flash events,is efficient in key-frame extraction and the extracted key frames can give a good visual summarization of the original motion video.

Key words: video indexing, key-frame, principal components analysis, clustering

摘要: 关键帧提取技术,对基于内容的视频检索有着重要的作用。为了能从不同类型的视频里有效的提取关键帧,提出了一种新的关键帧提取算法。首先通过主成分分析法提取视频特征信息,然后根据视频内容的复杂度自适应获得聚类数以及聚类中心,通过k均值聚类得到视频关键帧。实验表明该算法能消除闪光灯的干扰,有效地找出代表视频主要内容,尤其是目标运动信息的关键帧。

关键词: 视频检索, 关键帧, 主成分分析法, 聚类