Design of Image Classification System Based on Feature Extraction

Article Preview

Abstract:

Image classification is a kind of image data mining method to classify different targets based on different features reflected in image information. The paper designed a kind of image classification system based on feature selection, which utilize feature selection and feature weight to optimize the features and obtain features that can reflect essential of classification, so as to improve image classification accuracy. Meanwhile, the paper gave material implementation method of main modules of image classification system. Image classification experiment based on the system proves effectiveness of the designed system.

You might also be interested in these eBooks

Info:

Periodical:

Key Engineering Materials (Volumes 474-476)

Pages:

1859-1864

Citation:

Online since:

April 2011

Export:

Price:

Permissions CCC:

Permissions PLS:

Сopyright:

© 2011 Trans Tech Publications Ltd. All Rights Reserved

Share:

Citation:

[1] K. Mikolajczyk and C. Schmid: Scale & affine invariant interest point detectors, IJCV, vol. 1, 2004, pp.63-86.

Google Scholar

[2] D. Lowe: Distinctive image features from scale-invariant key points, IJCV. vol. 2, pp.91-110.

Google Scholar

[3] Nie Q. and Zhang S. Y.: Region-based image classification, Transaction of Beijing Institute of Technology, vol. 28, 2008, pp.885-889.

Google Scholar

[4] Hu B. B. and Yao M. H.: Image classification by using SVM, IJCV. vol. 1, 2010. pp.115-116.

Google Scholar

[5] Fu L., Sun H. F., Yang Y., Gu X. C. and Sun S. Z.: The Technnology of Image Classification on Bayesian Classifier, Journal of Changchun University of Science and Technology (Natural Science Edition), vol. 32, 2009, pp.132-134.

Google Scholar

[6] Wu H. and Yang K. J.: Kd- Tree-BasedMulti-Dimensional Indexing in the Database Application, Computer Applications, vol. 26, 2007, pp.37-39.

Google Scholar