计算机应用与软件2012，Vol.29Issue(5)：57-63,7.
基于MPEG-7的视频语义检索系统
AN MPEG-7 BASED VIDEO SEMANTIC RETRIEVAL SYSTEM
摘要
Abstract
A distributed video retrieval system that supports semantics, called XuNet, is designed and implemented. It uses an improved video semantic processing tool, which is based on IBM VideoAnnEx annotation tool and equipped with shot semantic graph annotation and natural language processing functions to semantically parse and annotate videos and generate MPEG-7 description files that contain semantic information. Afterwards XuNet builds distributed indices for videos' MPEG-7 files, at the meantime distributedly storing video files. The system provides abundant web query interfaces, including keywords semantic expansion query, semantic graph query and natural language query. As soon as a user submits his semantic query intention, XuNet quickly retrieves videos or segments according to his interests or even allows him to browse or play videos on demand. Owing to the distributed architecture, the whole system is highly scalable and supports mass video information indexing and retrieval.关键词
MPEG-7/语义/视频检索/Lucene/分布式Key words
MPEG-7/ Semantic/ Video retrieval/ Lucene/ Distributed system分类
信息技术与安全科学引用本文复制引用
郑烇,李少波,孙志军,袁婧,王嵩..基于MPEG-7的视频语义检索系统[J].计算机应用与软件,2012,29(5):57-63,7.基金项目
科技部支撑计划(2008BAH28B04) （2008BAH28B04）
安徽高校省级自然科学研究重点项目(KJ2009A152) （KJ2009A152）
