PERFORMANCE AND RELIABILITY FOR IRIS RECOGNITION BASED ON CORRECT SEGMENTATION

The recognition of Iris is regarded as the most dependable and accurate system of biometric identification so far. This system captures an individual's eye image in which the iris in the image is used for further normalization as well as segmentation to extract its feature. The iris recognition...

Full description

Saved in:
Bibliographic Details
Published in:Journal of Theoretical and Applied Information Technology 2016-11, Vol.93 (2), p.461-461
Main Authors: Jassim, Khider Nassif, Nseaf, Asama Kuder
Format: Article
Language:English
Subjects:
Accuracy
Biometric recognition systems
Biometrics
Eyes
Hough transforms
Image segmentation
Pupils
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The recognition of Iris is regarded as the most dependable and accurate system of biometric identification so far. This system captures an individual's eye image in which the iris in the image is used for further normalization as well as segmentation to extract its feature. The iris recognition systems performance relies heavily on the segmentation process. In fact, segmentation process is employed for localizing the accurate iris region in a certain portion of an eye and this must be correctly and accurately carried out to take out the eyelashes, reflection, eyelids and pupil noises found in the region of iris. In this study, Enhance Hough Transform (EHT) approach in the segmentation process will be used. The enhancement locates the pupil region of eye image by using threshold and Circle Hough Transform (CHT). Hence, the pupil parameter will capture the region of iris from the image of eye and then apply Hough Transform for locating outer boundary in less space search. This approach is found more effective in emphasizing the accuracy of iris segmentation. The segmented iris region is normalized so as to reduce the dimensional inconsistencies among the regions of iris through adopting the Model of Daugman's Rubber Sheet. The iris features were, then, encoded through convolving the normalized region of the iris with 1D Log-Gabor filters and phase quantizing the output so as to create a bit-wise biometric template. The distance of Hamming was selected as corresponding metric which provided the measure of a number of bits which did not match up among the iris templates. This proposed method is tested with the eye images obtained from MMU V1 iris database. The performance of such a proposed method showed that the accuracy of the iris recognition increased.
ISSN:1817-3195