Iris Segmentation in Visible Wavelength Environment
Abduljalil Radman, Kasmiran Jumari, Nasharuddin Zainal · Procedia Engineering · 2012
Iris segmentation is the difficult and essential step in iris recognition systems. The integrodifferential operator and Hough transform are widely used for segmenting iris texture of iris images. Both algorithms offer good segmentation accuracy but are computationally complex. However, a large portion of the computation time is consumed to identify initial centers of iris and pupil circular boundaries. This paper addresses simple solution for this problem by means of localizing the initial center of the pupil. The circular Gabor filter (CGF) is utilized to localize the initial pupil center. Performance studies have been evaluated using the UBIRIS.v1 iris image database. Results show that the proposed algorithm accelerates the segmentation process using the traditional integrodifferential operator and Hough transform, and simultaneously improves the segmentation accuracy.