Algorithm for Automated Iris Recognition Using Biorthogonal Wavelets

Algorithm for Automated Iris Recognition Using Biorthogonal Wavelets

Copyright: © 2014 |Pages: 7
DOI: 10.4018/978-1-4666-4896-8.ch012
OnDemand:
(Individual Chapters)
Available
$37.50
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

A comparative study of ability of two novel image retrieval algorithms to provide automated touch-free identification of persons by iris recognition is presented. Namely, applied biorthogonal wavelet methods and the SVD-Free Latent Semantic method are analyzed. Moreover, in case of the applied biorthogonal wavelet method, two approaches were tested and compared: the processing of the whole image and the processing of the image converted to a vector. The point is that different methods for getting rid of the noise were successfully applied in both cases. Numerical experiments on a real biometric database indicate feasibility of the presented approach as an automated iris recognition tool without special image pre-processing.
Chapter Preview
Top

3. Biorthogonal Wavelets

The biorthogonal wavelets are used in this study (Gonzales, 2004), (Prasad, 1997) as a tool for iris recognition problem. Two techniques are applied. Two-dimensional and one dimensional biorthogonal wavelet filters are used.

The first technique (Xu, 1994), (Chang, 2000) implies procedure with an image converted to gray image. Even though the processing of Red component of images has shown the best results, in this case the procedure of gray images appeared to give superior results. Every image is processed as a two-dimensional m×n matrix image. The two-dimensional biorthogonal wavelets are applied to gray images.

Complete Chapter List

Search this Book:
Reset