A robust algorithm for iris localization based on radial symmetry and circular integro differential operator

Locating an iris is important in an iris recognition system. Previous studies have shown good performance when an iris image is clear. However, if an iris image is occluded by eyelids, eyelashes and reflection, the accuracy and robustness of localization will decrease significantly. This paper propo...

Full description

Saved in:
Bibliographic Details
Main Authors: Zuhong Zhou, Peng Yao, Zhenquan Zhuang, Jun Li
Format: Conference Proceeding
Language:English
Subjects:
Eyelashes
Eyelids
Iris
Iris recognition
Pixel
Robustness
Transforms
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Locating an iris is important in an iris recognition system. Previous studies have shown good performance when an iris image is clear. However, if an iris image is occluded by eyelids, eyelashes and reflection, the accuracy and robustness of localization will decrease significantly. This paper proposed a new algorithm to locate an iris in a coarse-to-fine manner: firstly, a modified radial symmetry transform was used to find the possible pupil region; secondly, an enhanced circular integro operator was adopted to obtain the precise iris location. Experimental results on CASIA V3.0 iris database showed that proposed algorithm has a more accurate and robust performance than the previous methods, especially when occlusion of eyelids, eyelashes and reflection occurred. A real-time system was able to be achieved by the coarse-to-fine strategy as well.
ISSN:2156-2318
2158-2297
DOI:10.1109/ICIEA.2011.5975873