Best Wavelet Function for Face Recognition Using Multi-Level Decomposition
The selection of appropriate wavelets is an important target for any application. In this paper, wavelets functions are examined in order to choose the best wavelet for face classification process and for finding the optimal number of levels of decomposition. Seven wavelet functions namely Symelt, D...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Published: |
2011
|
| Subjects: | |
| Online Access: | http://eprints.utp.edu.my/7141/1/A7_PresentSchedule_071011.pdf http://eprints.utp.edu.my/7141/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The selection of appropriate wavelets is an important target for any application. In this paper, wavelets functions are examined in order to choose the best wavelet for face classification process and for finding the optimal number of levels of decomposition. Seven wavelet functions namely Symelt, Daubechig, Coiflets, Mayer Discrete, Biorthogonal, Reverse Biorthogonal and Haar were tested with different number of decomposition levels and different number of biggest coefficients is selected to reduce the huge feature dimension, and then the Euclidean Distance Method (EDM) was used for classification process. Also a statistical method has been proposed to produce new metric of features coefficients, the experiments brought about 40% improvements in comparison to the method that accounts the biggest coefficients from the four levels of decompositions. The experiments have been performed on Olivetti Research Laboratory database (ORL) and Yale University database (YALE). The result showed the effect of wavelets proprieties on classification process and the Symelt wavelets are the optimum wavelets for the face classification with four levels. |
|---|