基于内容的静态语义概念视频检索方法研究

Research of Content-Based Video Retrieval on Static Semantics Concept

本文针对互联网视频检索技术的发展,阐述了目前主流视频搜索引擎的技术现状,分析了互联网视频检索的关键技术,特别是对于视频特征的提取技术。本文的创新点是提出了一种通用的基于内容的静态语义视频检索方法,该方法可以弥补基于文本视频检索的有关不足,并且在TRECVID的视频概念检索数据的静态语义概念中得到验证,运行稳定。

This paper does the research on the main technology status of the current mainstream Internet video search engine, to do a detailed analysis of the key technologies of the Internet Video Retrieval. Especially, video feature extraction is detailed presents in this paper. And this paper presents a general method of content-based video retrieval on static semantics concept. This method can make up the related deficiencies of traditional text-based video retrieval. Meanwhile, this method has a great performance on static semantics content of the TRECVID database.