In this paper we present results that expand on the Afghan Girl study by Daugman using gray-scale photographs of 4 subjects taken every year for 40 years: The Brown Sisters photographic project by Nicholas Nixon. We extracted iris images from the photographs and generated match scores using two commercial iris recognition algorithms. We developed criteria for making objective decisions about the utility of the extracted iris images and restricted the analysis of the mated pairs to those satisfying the criteria. The resulting match scores demonstrate matches at thresholds that correspond to FMR of 0.01 or better over periods of up to 9 years. These are the initial results; optimization of the process may improve the results. The Brown Sisters images used in this work are ©Nicholas Nixon, and courtesy Fraenkel Gallery, San Francisco.
Technical Note (NIST TN) - 2098
iris recognition, biometrics, Nicholas Nixon, Brown Sisters