基于增量主成份分析的在线视频镜头边界检测  

Online Shot Boundary Detection Based on Incremental Principal Component Analysis

作　　者：朱庆生[1] 孟贵超[1] 葛垚[1] 

机构地区：[1]重庆大学计算机学院,重庆400044

出　　处：《计算机工程与科学》2012年第1期94-97,共4页Computer Engineering & Science

基　　金：国家863计划资助项目(2006AA10Z233);国家自然科学基金资助项目(60773082)

摘　　要：在基于内容的视频检索系统中,镜头边界检测是极其关键的一步。而在线检测镜头边界也是亟需解决的问题,所以提出利用增量主成份分析(IPCA)方法,对视频进行实时分析和在线镜头边界检测。IPCA方法允许样本逐步输入,这样正符合在线视频的获取方式,所以利用IPCA可以实现镜头边界的在线检测。但是,IPCA也存在存储空间逐步增大的缺陷,针对此问题也提出了解决方案。实验证明该方法可以取得满意的效果。Shot boundary detection is the key point of content-based video retrieval. In order to implement online detection, a method based on incremental principal component analysis （IPCA） is proposed to analyze video in a real-time way in this paper, and then to get the shot boundary. Its advantage is detecting shot boundary and extracting Abstract video without acquiring the whole video in advance, in virtue of which it makes the online detection come true. But it needs large memory, which can be solved by the way proposed. The experiment proves it has good performance.

关 键 词：基于内容的视频检索 镜头边界检测 增量主成份分析 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]