A field study of the accuracy and reliability of a biometric iris recognition system

Abstract

Purpose: The iris of the eye appears to satisfy the criteria for a good anatomical characteristic for use in a biometric system. The purpose of this study was to evaluate a biometric iris recognition system: Mobile-Eyes™.

Methods: The enrollment, verification, and identification applications were evaluated in a field study for accuracy and reliability using both irises of 277 subjects. Independent variables included a wide range of subject demographics, ambient light, and ambient temperature. A sub-set of 35 subjects had alcohol-induced nystagmus. There were 2710 identification and verification attempts, which resulted in 1,501,340 and 5540 iris comparisons respectively.

Results: In this study, the system successfully enrolled all subjects on the first attempt. All 277 subjects were successfully verified and identified on the first day of enrollment. None of the current or prior eye conditions prevented enrollment, verification, or identification. All 35 subjects with alcohol-induced nystagmus were successfully verified and identified. There were no false verifications or false identifications. Two conditions were identified that potentially could circumvent the use of iris recognitions systems in general.

Conclusions: The Mobile-Eyes™ iris recognition system exhibited accurate and reliable enrollment, verification, and identification applications in this study. It may have special applications in subjects with nystagmus.