ThAT9.20 - Face Recognition Using Anisotropic Dual-Tree...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Face Recognition Using Anisotropic Dual-Tree Complex Wavelet Packets * Yigang Peng, Xudong Xie, Wenli Xu and Qionghai Dai Department of Automation, Tsinghua University, Beijing 100084, China pyg07@mails.tsinghua.edu.cn, { xdxie,xuwl,qhdai } @mail.tsinghua.edu.cn Abstract In this paper, we propose a novel face recognition method based on anisotropic dual-tree complex wavelet packets(ADT-CWP). 2-D dual-tree complex wavelet transform(DT-CWT) provides a geometrically oriented decomposition for image representation as well as shift invariance. By applying anisotropic wavelet packet decomposition on DT-CWT further, ADT-CWP can be used to extract facial features better, which turns out to benefit for face recognition. With adaptively assigning different weights to different wavelet subbands, consis- tent best performances can be obtained based on dif- ferent face databases which are under different condi- tions, such as varying illuminations and expressions, compared to PCA and other face recognition methods, especially Gabor-based method. Furthermore, in addi- tion to the consistent and promising classification per- formances, our proposed ADT-CWP-based method has a really low computational complexity. 1. Introduction Face recognition has always been a hot research topic within the last several years. It has a wide range of applications such as security systems, credit card veri- fication, scene surveillance including commercial and law enforcement applications [1]. Numbers of algo- rithms have been proposed. However, reliable tech- niques have proven elusive because their performances vary due to image variations caused by illumination conditions, facial expressions, poses or perspectives and other factors. In this paper, we propose an efficient face recognition method based on anisotropic dual-tree com- plex wavelet packets(ADT-CWP), using a single im- age per person for training. In our method, an ADT- CWP decomposition structure is achieved by perform- ing anisotropic wavelet packet decomposition on an av- * This work is supported by the Distinguished Young Scholars of NSFC (Grant No. 60525111), the Joint Fund for Overseas Chinese Young Scholars of NSFC (Grant No. 60528004), and the 863 Pro- gram (Grant No. 2007AA01Z332). erage face. Then this decomposition structure is ap- plied to every face image for facial features extraction in training and testing procedures. The magnitudes of different wavelet subbands coefficients are normalized and modified by multiplying with the standard devia- tion of the corresponding subbands obtained from the average face for better feature representation. The Eu- clidean distance metric is employed on these modified coefficients for classification. We evaluate the perfor- mance of our proposed method for face recognition with the use of four different databases, and consistent and promising results can be obtained, which reveals that our method can improve the recognition performances in all conditions at a cost of low computational com-...
View Full Document

This note was uploaded on 05/28/2010 for the course EE EE564 taught by Professor Runyiyu during the Spring '10 term at Eastern Mediterranean University.

Page1 / 4

ThAT9.20 - Face Recognition Using Anisotropic Dual-Tree...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online