Ocular Biometric System Focused on Iris Localization and Embedded Matching Algorithm
Eri Prasetyo Wibowo, Karmila Sari, Wisnu Sukma Maulana, Sarifudin Madenda · International Journal of Computer and Electrical Engineering · 2010
This research aims to develop previous researches which published in the ICSPS 2009 , that were successful to perform iris localization. The indicator values perform well for doing iris localization in comparison with two previous works of Daugman's and Wildes' work. Even succeeded in iris localization, previous researches still has some weaknesses. The worst weakness about to the matching between two iris, even when applied with Hamming and Euclidean distance to perform the code matching of iris.the average similarity for matching two iris image taken from same iris is just 30 to 45 percents. Hereby, research to propose a research in ocular biometric system to identify someone by using iris part. This research is still focused on how to perform iris recognition by doing the normalization and feature extraction to get better iris localization. After iris localization step is performed, then it is continued by extracting the iris feature to get the iris code. The last step is to perform iris code matching embedded with shift bit algorithm in hamming distance to get the better result. Index Terms—Algorithm, bit shift, code, iris, matching.