A New Method for Iris Localization

Iris localization is one of the most important steps in iris recognition system and determines the accuracy of matching. In this paper, a new method for iris localization is proposed. Firstly, we adopt edge points detecting to find out the iris' inner boundary. Next, using the SUSAN(smallest unvalued segment assimilating nucleus) edge detector, the coarse outer boundary will been found, and then obtain four edge points through four certain directions. Finally, we compare the four edge points and find out the outer boundary. The algorithm is proved to be effective through our experiments on 2400 iris images from CASIA (The Institute of Automation, Chinese Academy of Sciences) database.

[1]  Lin Zhonghua,et al.  A Novel Iris Localization Algorithm Based on the Gray Distributions of Eye Images , 2005, 2005 IEEE Engineering in Medicine and Biology 27th Annual Conference.

[2]  Tim J. Dennis,et al.  An adaptive implementation of the SUSAN method for image edge and feature detection , 1997, Proceedings of International Conference on Image Processing.

[3]  Richard P. Wildes,et al.  Iris recognition: an emerging biometric technology , 1997, Proc. IEEE.

[4]  Mei Xie,et al.  Research on iris image preprocessing algorithm , 2005, IEEE International Symposium on Communications and Information Technology, 2005. ISCIT 2005..

[5]  Stephen M. Smith,et al.  SUSAN—A New Approach to Low Level Image Processing , 1997, International Journal of Computer Vision.

[6]  John Daugman,et al.  High Confidence Visual Recognition of Persons by a Test of Statistical Independence , 1993, IEEE Trans. Pattern Anal. Mach. Intell..