Integrated Video and Text for Content-based Access to Video Databases

  • Authors:

  • Haitao Jiang;Danilo Montesi;Ahmed K. Elmagarmid

  • Affiliations:

  • Computer Science Department, Purdue University, West Lafayette, IN 47907 USA. jiang@cs.purdue.edu;Dipartimento di Scienze dell‘Informazione, Universita‘ di Milano Via Comelico 39/41 20135 Milano, Italy. montesi@dsi.unimi.it;Computer Science Department, Purdue University, West Lafayette, IN 47907 USA. ake@cs.purdue.edu

  • Venue:
  • Multimedia Tools and Applications
  • Year:
  • 1999

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper introduces a new approach to realize video databases. Theapproach consists of a VideoText data model based on free textannotations associated with logical video segments and a correspondingquery language. Traditional database techniques are inadequate forexploiting queries on unstructured data such as video, supportingtemporal queries, and ranking query results according to theirrelevance to the query. In this paper, we propose to use informationretrieval techniques to provide such features and to extend the querylanguage to accommodate interval queries that are particularly suitedto video data. Algorithms are provided to show how user queries areevaluated. Finally, a generic and modular video database architecturewhich is based on VideoText data model is described.