Development of automated iris identification system prototype
Increasing threats in security has led identification technology to the era of biometrics identification. With identification in forms of physical traits that are distinctive to each individual, biometrics offer enhanced security over traditional identification systems. Regarded as one of the most accurate personal identification, iris is highly unique between individuals, leading to many active studies in this field (Daugman, 2004; Wildes, 1997). The Agency for the Assessment and Application of Technology (Badan Pengkajian dan Penerapan Teknologi; BPPT) initiate the research on Multimodal Biometrics identification system, where iris identification is an integral part. This research is conducted as an initial study and implementation of the algorithm in form of a prototype. The functionality of the prototype covers iris localization, normalization, encoding and matching. A novel approach is proposed to accurately localized iris and the normalization is based on the method proposed by Miyazawa, et al., whereas the subsequent processes refer to the scheme by Daugman (Daugman, 2004; Miyazawa, et al., 2006). Based on the experiments, the success rates for iris localization is 99.56% and with resulting EER of 1600 comparisons at 0.40.
B01169 | (wh) | Available |
No other version available