VECTOR-BASED COLOUR IMAGE RETRIEVAL

Khor, Siak Wang and Abdullah, Muhamad Taufik (2007) VECTOR-BASED COLOUR IMAGE RETRIEVAL. In: 2nd International Conference on Informatics (Informatics 2007), Hilton Petaling Jaya Hotel, Petaling Jaya, Selangor, Malaysia.

[img] PDF
IMS.pdf

Download (1MB)
Official URL: http://dspace.fsktm.um.edu.my/handle/1812/401

Abstract

This paper describes a novel approach for retrieving images which are chromatically similar to a query image. The technique makes use of vector model from the literatures of information retrieval for image similarity match. With a properly devised colour space discretisation technique, collections of pixels having similar colour content are identified, grouped and sorted according to the proportional spread, in term of total pixels in a given image. Weights are assigned to each identified groups. The sorted data will then be stored in an ASCII file, ready for the image retrieval process. The proposed technique will be benchmarked against the histogram-based approach, which is a common approach for colour image retrieval, in term of retrieval accuracy. A standard dataset that currently contains 1338 colour images, known as Uncompressed Colour Image Database (UCID), is used for the benchmarking purpose. The proposed technique has shown an improved performance in term of retrieval accuracy over the traditional histogram-based approach.

Item Type: Conference or Workshop Item (Paper)
Subjects: Z Bibliography. Library Science. Information Resources > Z665 Library Science. Information Science
Divisions: Faculty of Engineering, Science and Mathematics > School of Electronics and Computer Science
Depositing User: MS SITI NUR ATIKAH MOHAMAD RUSDI
Date Deposited: 15 Jul 2013 07:24
Last Modified: 15 Jul 2013 07:24
URI: http://repository.um.edu.my/id/eprint/308

Actions (login required)

View Item View Item