计算机应用 ›› 2010, Vol. 30 ›› Issue (3): 835-837.

• 典型应用 • 上一篇    下一篇

基于本体的视频语义检索系统研究

徐峰,郑烇   

  1. 中国科学技术大学
  • 收稿日期:2009-09-17 修回日期:2009-11-12 发布日期:2010-03-14 出版日期:2010-03-01
  • 通讯作者: 徐峰
  • 基金资助:
    语义视频检索

Research of video semantic retrieval system based on ontology

  • Received:2009-09-17 Revised:2009-11-12 Online:2010-03-14 Published:2010-03-01
  • Contact: feng xu

摘要: 在语义层次上检索视频内容,可以突破“语义鸿沟”,提高视频内容的利用效率。利用本体的标注和推理能力,研究视频语义检索,充分挖掘视频内容的结构和语义信息,构建层次化的语义索引,能极大提高系统的语义检索能力。该视频语义检索系统(OVSR)的本体架构集成了领域本体、视频本体和核心本体,具有很强的扩展和交互能力。主要论述OVSR的本体架构,视频语义模型和索引模型,研究OVSR的查询重写算法以及本体推理算法。

关键词: 本体, 语义索引, 推理能力, 查询重写, 索引模型

Abstract: Retrieving video content in the semantic level can break "semantic gap" and increase the utilization efficiency of video content. The authors made use of the annotation and reasoning ability of ontology, studied video semantic retrieval, fully mined the structural and semantic information of the video content, and built the hierarchical semantic indexing, which can greatly strengthen the system's semantic retrieval ability. The ontology structure of Video Semantic Retrieval System (OVSR) integrated domain ontology, video ontology and core ontology, and has a strong expanding and interoperability ability. The authors mainly discussed OVSR's ontology structure, video semantic model and indexing model, and studied the user's searching rewriting algorithm and ontology reasoning algorithm.

Key words: ontology, semantic retrieval, reasoning ability, searching rewriting, indexing model