Iris Biometrics: Template Protection and Advanced Comparators

Iris recognition is gaining popularity as a robust and reliable biometric technology. The intricate structure of the iris constitutes a powerful biometric modality utilized by iris recognition algorithms to extract biometric templates. Proposed approaches report recognition rates above 99% and equal error rates less than 1% on diverse data sets. Providing high accuracy iris recognition appears to be well suitable for access control systems managing large-scale user databases. Form a privacy perspective most concerns against the common use of biometrics arise from the storage and misuse of biometric data, e.g. tracking persons without consent. Biometric cryptosystems and cancelable biometrics represent emerging technologies of biometric template protection addressing these concerns and improving public confidence and acceptance of biometric systems. In order to protect templates from infiltration, e.g. based on brute-force attacks, underlying biometric features are required to exhibit sufficient entropy, i.e. iris represents the biometric modality of choice for high security authentication based on template protection technologies. Most publications regarding iris recognition aim at extracting discriminative biometric templates while only few, usually trivial, comparison techniques, e.g. fractional Hamming distance, have been proposed. Advanced iris-biometric comparators have received only little consideration, i.e. potential improvements in the comparison stage are frequently neglected. In this cumulative thesis iris-biometric template protection and advanced comparators are investigated. Based on detailed descriptions of published approaches in both research subareas an overview and discussion, including an experimental study, are presented.