Title: Local-feature-based image retrieval with weighted relevance feedback

Authors: Rudra Narayan Hota, Shahanaz Syed, P. Radha Krishna

Addresses: SET Labs, Infosys Technologies Limited, Hyderabad 500-032, India. ' SET Labs, Infosys Technologies Limited, Hyderabad 500-032, India. ' SET Labs, Infosys Technologies Limited, Hyderabad 500-032, India

Abstract: Accurate and fast retrieval of relevant images is a challenging task mainly due to the limitation in understanding hidden knowledge in images, known as semantic gap. In this work, we propose a novel approach which incorporates local feature representation for retrieval of grey and colour images from an archive with user intervention. We used histogram features, which are computationally efficient, hence resulting in quick image retrieval. The computed image feature vectors are used for similarity matching with weighted feed-backed image retrieval. We experimented both on publicly available and annotated image data sets to illustrate the effectiveness of our approach.

Keywords: relevance; weighted feedback; local features; extraction; colour; texture; image retrieval; accuracy; speed; fast retrieval; hidden knowledge; semantic gaps; feature representation; grey images; archives; user intervention; histograms; computed images; vectors; similarity matching; public data sets; annotated data sets; annotations; multimedia; XML streams; querying; extensible markup language; data mining; business intelligence.

DOI: 10.1504/IJBIDM.2010.036124

International Journal of Business Intelligence and Data Mining, 2010 Vol.5 No.4, pp.353 - 369

Published online: 22 Oct 2010 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article