In this study, we proposed a quick and accurate algorithm for content-based image classification. The proposed method is also used to retrieve similar images from databases. In this paper color and texture information are used to represent image features. The basic idea is to extract color information about global and local features of images. A global color feature is extracted by an RGB model. While, a local color feature is extracted by an HSV model. In the case of a local feature, if it cannot be classified, the result is inaccurate retrieval. A GA (genetic algorithm) is used to extract local features which can be classified. Local features extracted by a GA are optimal representative features. In the experiment, the accuracy of image classification is measured using the proposed algorithm. Also, we compared the previous algorithm with the proposed algorithm in terms of image classification performance. As a result, the proposed algorithm showed higher performance in terms of accuracy.
Bibliographical noteFunding Information:
This research was supported by the MKE (Ministry of Knowledge Economy), Korea, under the ITRC (Information Technology Research Center) support program supervised by the IITA (Institute of Information Technology Advancement) (IITA-2008-(C1090-0801-0025)).
This work was supported by the Brain Korea 21 Project in 2008.
Copyright 2009 Elsevier B.V., All rights reserved.
- Feature information
- Genetic algorithm
- Image retrieval
- Support vector machine
ASJC Scopus subject areas
- General Engineering
- Computer Science Applications
- Artificial Intelligence