Distributional clustering for efficient content-based retrieval of images and video

11 November 2002

conference paper
Published by Institute of Electrical and Electronics Engineers (IEEE)

Vol. 1, 81-84 vol.1
https://doi.org/10.1109/icip.2000.900897

Abstract

We present an approach to clustering images for efficient retrieval using relative entropy. We start with the assumption that visual features are represented by probability densities and develop clustering algorithms for probability densities (for example, normalized histograms are crude approximations of probability densities). These clustering algorithms are then used for efficient retrieval of images and video.

Keywords

This publication has 7 references indexed in Scilit:

VideoBook: an experiment in characterization of video
Published by Institute of Electrical and Electronics Engineers (IEEE) ,2002
A Bayesian framework for semantic content characterization
Published by Institute of Electrical and Electronics Engineers (IEEE) ,2002
Content analysis of video using principal components
Published by Institute of Electrical and Electronics Engineers (IEEE) ,2002
Content-based search of video using color, texture, and motion
Published by Institute of Electrical and Electronics Engineers (IEEE) ,2002
Photobook: Content-based manipulation of image databases
International Journal of Computer Vision, 1996
Query by image and video content: the QBIC system
Computer, 1995
Vector quantization by deterministic annealing
IEEE Transactions on Information Theory, 1992