One major factor of finger vein authentication performance variability is the difference in the recognizability of different users. It may be effective to enhance the authentication performance by improving the recognizability of each user. In this paper, we determine a specific threshold for each user to carry out identity authentication, comparing to the traditional common threshold. The user-specific threshold is calculated according to user's impostor matching scores. We compare the authentication performances of the common threshold and the user-specific threshold respectively in the verification mode and the recognition mode. The experiments perform on our self-built finger vein database, and experimental results show that the user-specific threshold possesses promising advantage in finger vein authentication. © Springer International Publishing 2013.
CITATION STYLE
Yang, L., Yang, G., Yin, Y., & Xiao, R. (2013). Exploring user-specific threshold in finger vein authentication. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8232 LNCS, pp. 306–313). https://doi.org/10.1007/978-3-319-02961-0_38
Mendeley helps you to discover research relevant for your work.