Multimedia index and retrieval under MARC framework

  • Authors:

  • Wenying Wendy Gao;Jesse S. Jin

  • Affiliations:

  • School of Computer Science and Engineering, University of New South Wales, Sydney, Australia;School of Computer Science and Engineering, University of New South Wales, Sydney, Australia and School of Information Technologies, University of Sydney, NSW, Australia

  • Venue:
  • VIP '02 Selected papers from the 2002 Pan-Sydney workshop on Visualisation - Volume 22
  • Year:
  • 2003

Quantified Score

Hi-index 0.01

Visualization

Abstract

Information retrieval is the indexing and retrieving of data that match criteria specified by users. In today's age, the proliferation of multimedia information makes it necessary to revolutionise traditional type of information retrieval systems. Content Based Image Retrieval (CBIR) is a research area attempting to address this issue. It is primarily concerned with using content-based information such as visual features to specify search criteria and perform query on image collections. A number of different focuses exist in CBIR research, ranging from research on feature extraction techniques, different indexing schemes to collection organisation. In a CBIR system, like in any information retrieval system, in order to achieve effective indexing and retrieving of data, it is very important to have a cataloguing framework that enables easy maintenance of collections, otherwise it would not be possible to put the system into practical use.CBIR-VU is a content-based image retrieval system developed in the VIP Lab, University of NSW. This system has the problem of lacking a well-defined cataloguing framework, which severely restricts the system domain, and renders it impractical for real use. During a thesis project this year, a catalogue based on an internationally recognised cataloguing framework - MARC 21 format - has been developed for CBIR-VU.This catalogue enables effective organisation of many different medium of information, ranging from multimedia information to traditional material, and it is able to provide a solid backbone to new generation information retrieval systems like CBIR-VU.