Biometric features are known to change over time, presenting a challenge for their use in identity management systems. Viewed as an instrumentation and measurement problem, these changes represent a potential source of measurement or calibration error that needs to be addressed at the system level in order to guarantee performance over the lifetime of the system. In this paper, we develop a novel metric, biometric permanence, to characterize the stability of biometric features. First, we define permanence in terms of the change in a false nonmatch rate (FNMR) over a repeated sequence of enrollment and verification events for a given population. However, since changes in the FNMR are expected to be small, any variability in the biometric capture over time will camouflage the changes of interest. To address this issue, we propose a robust methodology that can isolate the visit-to-visit variability and substantially improve the estimation. We develop and characterize a heuristic statistical model for a biometric capture system, and apply it to a large data set of fingerprint biometrics collected over a period of seven years on a variety of commercially available capture devices. We discuss how this methodology can be used to isolate the effect of biometric template aging and to develop system-level strategies for dealing with it.

, , , , , , , , , , , , ,
IEEE Transactions on Instrumentation and Measurement
Department of Systems and Computer Engineering

Harvey, J. (John), Campbell, J. (John), & Adler, A. (2018). Characterization of Biometric Template Aging in a Multiyear, Multivendor Longitudinal Fingerprint Matching Study. IEEE Transactions on Instrumentation and Measurement. doi:10.1109/TIM.2018.2861998