Fingerprint dynamics: A novel biometrics for personal authentication

Biometric has emerged as an eminent solution to different security and privacy issue of access control applications. Recently many novel physiological and behavioral characteristics have been used as biometric identifiers. In this paper, we propose a novel approach based on behavioral characteristics, namely Fingerprint Dynamics. The technique is validated as a standalone identity verification system. For this a dataset from 19 individuals is collected in form of multi-instance fingerprint scan events. The behavioral characteristics associated with fingerprint scan events is analyzed. An acquisition setup (hardware) is also devised to capture the user data. A neural network based classifier is used for the purpose of identity verification. Aim of this paper is to assess the potential of fingerprint dynamics to uniquely epitomize a person based on his behavioral characteristics. Verification experiment in presence of zero-effort imposter attack confirms the discriminating capability of proposed technique.