摘要
就海量视频数据进行标引的方法进行了阐述,对元数据、Dublin Core、OAIS进行了分析.通过研究这些技术在视频挖掘中所具有的优势,提出了一种海量视频数据标引平台的体系结构并实现了相关的功能模块,并对视频数据的搜索提出了一个基于标引的解决方法.实验结果证明,该平台可为互联网视频搜索的发展提供更加快捷、方便、准确的标引和检索模式,有效降低了用户获取相关视频数据的时间.
This paper describes indexing methods for massive video data. It analyzes metadata, Dublin Core, and the open archival information system （OAIS） in detail. To apply these technological advances to video mining, we suggested a platform for massive video data indexing and the relevant functional modules were established. Furthermore, a solution for video searches based on indexing was proposed. It was proven by experiments that this platform, with its more detailed and convenient indexing approach, would be a great help to the development of Internet video searches by effectively saving users＇ time and energy in the search for valuable data.
出处
《智能系统学报》
2009年第2期107-111,共5页
CAAI Transactions on Intelligent Systems
基金
国家"863"计划资助项目(2009AA01Z143)
铁道部-清华大学科研计划资助项目(J2008X009)
