Automatic Iris Segmentation Based on Local Areas

Author

Xu, Guangzhu ; Zhang, Zaifeng ; Ma, Yide

Author_Institution

Sch. of Inf. Sci. & Eng., Lanzhou Univ.

Volume

4

fYear

0

fDate

0-0 0

Firstpage

505

Lastpage

508

Abstract

In this paper a novel and robust method for automatic iris segmentation based on local areas is described. Such method is composed of three main parts, (a) Find the local rectangle region which has the minimum intensity mean and extend it to locate pupil, (b) Select two small local sector areas including the outer boundaries of iris to locate outer iris, (c) Translate the iris from polar coordinates into Cartesian coordinates and normalize it to fixed size to compensate the stretching of the iris texture as the pupil changes in size and remove the non-concentricity of the iris and the pupil. The method was implemented using CASIA iris image databases. The experimental results show that the proposed method has an encouraging result with an overall accuracy of 98.42%

Keywords

eye; geometry; image segmentation; image texture; CASIA iris image database; Cartesian coordinates; automatic iris segmentation; iris texture; local rectangle region; polar coordinates; Authentication; Biometrics; Eyelashes; Eyelids; Feature extraction; Fingerprint recognition; Image databases; Image segmentation; Iris recognition; Robustness;

fLanguage

English

Publisher

ieee

Conference_Titel

Pattern Recognition, 2006. ICPR 2006. 18th International Conference on

Conference_Location

Hong Kong

ISSN

1051-4651

Print_ISBN

0-7695-2521-0

Type

conf

DOI

10.1109/ICPR.2006.300

Filename

1699889