PISARO: Perceptual Colour and Texture Queries using Stackable Mosaics

  • Authors:

  • Matthew Seaborn;Lee Hepplewhite;John Stonham

  • Affiliations:

  • Brunel University;Brunel University;Brunel University

  • Venue:
  • ICMCS '99 Proceedings of the IEEE International Conference on Multimedia Computing and Systems - Volume 2
  • Year:
  • 1999

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper introduces PISARO (Perceptual Image Similarity And Retrieval Online), a web based image retrieval system designed for combining, evaluating and comparing Content Based Image Retrieval (CBIR) techniques. The need for a query interface which allows the user to understand both query and subsequent retrievals is identified and implemented together with a stackable mosaic approach to query design which gives the benefit of pre-calculated features whilst allowing the user to paint queries. Several colour and texture based feature extraction methods are evaluated for CBIR including a new Perceptual Colour Histogram. Finally, the suitability of the stackable mosaic approach is demonstrated using colour and texture query combinations.