Spectroscopic Approach for Aliveness Detection in Biometrics Authentication

This paper proposes the use of spectroscopic approach as a means to enhance existing biometrics technology in order to prevent spoofing. It shows how biometrics authentication systems can be enhanced with a spectroscopic method in a multi-factor manner such that a person´s unique ´spectral signatures´ or ´spectral factors´ are recorded and compared in addition to a non-spectroscopic biometric signature to reduce the likelihood of imposter getting authenticated. This paper also shows how to extract the ´spectral factors´ (i.e., reflectance values that corresponds to physiological components like melanin, hemoglobin, arterial and venous blood) from reflectance spectra of fingers and use cluster analysis in order to monitor and check aliveness and authenticity of the person during a fingerprint verification process.