Performance and computational burden of histogram based color image retrieval techniques

Authors:
K. Konstantinidis;I. Andreadis
Affiliations:
(Correspd. Tel.: +30 541079566/ Fax: +30 541079564/ E-mail: konkonst@ee.duth.gr) Lab. of Electronics, Section of Electronics and Info. Sys. Technol., Dept. of Electrical and Computer Eng., Democri ...;Laboratory of Electronics, Section of Electronics and Information Systems Technology, Department of Electrical and Computer Engineering, Democritus University of Thrace, GR-67100 Xanthi, Greece
Venue:
Journal of Computational Methods in Sciences and Engineering
Year:
2005

Citing 3
Cited 2

Introduction to statistical pattern recognition (2nd ed.)

Introduction to statistical pattern recognition (2nd ed.)
Color indexing

International Journal of Computer Vision
Visual information retrieval

Visual information retrieval

An efficient peer to peer image retrieval technique using content addressable networks

SETN'06 Proceedings of the 4th Helenic conference on Advances in Artificial Intelligence
Content-based image retrieval using OWA fuzzy linking histogram

Journal of Intelligent & Fuzzy Systems: Applications in Engineering and Technology - Computational intelligence models for image processing and information reasoning

Quantified Score

Hi-index	0.00

Visualization

Abstract

Due to the fact that image libraries have largely become overcrowded, the need of improved methods for retrieving color images from such libraries has increased immensely. Color-based Image Retrieval (CBIR) is a collection of techniques for retrieving images on the basis of color. An efficient tool, which is widely used in CBIR, is image color histograms. In this paper we present a comparison among the methods used to compare global and local histograms in the color spaces of RGB, HSV, L*a*b* and LCH. Moreover, the computational burden of all the systems carried out was compared and discussed both in terms of CPU time and number of computations. One of the key results from the experiments with global histograms was that the main computational burden was produced by the color space transforms rather than the comparative techniques. All the comparisons of the featured methods were performed through simulations under the environment of the MathWorks' Matlab package.