By Topic

Face recognition using multi-scale PCA and Support Vector Machine

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$31 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

2 Author(s)
Guoyun Zhang ; Dept. of Phys. & Electron. Inf., Hunan Inst. of Sci. & Technol., Yueyang ; Jing Zhang

Based on Gabor wavelets, a novel multi-scale principal component analysis and support vector machine algorithm (MsPCA-SVM) for face recognition is proposed in this paper. Firstly, the Gabor wavelets transformation results including five scales and eight directions are calculated and 40 feature matrices which are reconstructed with the same scale and the same direction transform results of the different face images are obtained. Secondly, the dimensionality reduction and denoised technique with PCA are applied to form the new training samples. Finally, 40 SVMs classifiers are constructed and the vote decision strategy is used to determine the recognition results. The experimental results show that the proposed method expands the selectable range of the cumulative variance contribution rate in PCA method and the influence of the SVMs kernel parameters on the recognition rate is small. So, the SVMs kernel parameters are easy to select. Furthermore, the difficult problem to select the kernel parameters has been settled to a certain degree. In the meantime, the ideal recognition rate is obtained.

Published in:

Intelligent Control and Automation, 2008. WCICA 2008. 7th World Congress on

Date of Conference:

25-27 June 2008