A New Feature Extraction for Iris Identification Using Scale-Space Filtering Technique

Jinil HONG
Woo Suk YANG
Dongmin KIM
Young-Ju KIM

IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences   Vol.E87-A    No.12    pp.3404-3408
Publication Date: 2004/12/01
Online ISSN: 
Print ISSN: 0916-8508
Type of Manuscript: PAPER
Category: Image
iris,  human identification,  biometrics,  scale-space filtering,  

Full Text: PDF>>
Buy this Article

In this paper, we introduce a new technology to extract the unique features from an iris image, which uses scale-space filtering. Resulting iris code can be used to develop a system for rapid and automatic human identification with high reliability and confidence levels. First, an iris part is separated from the whole image and the radius and center of the iris are evaluated. Next, the regions that have a high possibility of being noise are discriminated and the features presented in the highly detailed pattern are then extracted. In order to conserve the original signal while minimizing the effect of noise, scale-space filtering is applied. Experiments are performed using a set of 272 iris images taken from 18 persons. Test results show that the iris feature patterns of different persons are clearly discriminated from those of the same person.