Multimodal Biometric Authentication in IoT: Single Camera Case Study

Abstract

This paper presents an approach to multimodal biometric authentication using face and iris biometric traits. Having in mind that variety of devices, such as laptops, smartphones and tablets have a high quality camera built in, it is possible to obtain images of iris and face simultaneously. By combining geometric and photometric techniques, such as fiducial point localization and Gabor filtering, features are extracted and biometric templates are generated. Generated templates are further used to identify and authenticate the user on any device which he is allowed to use, thus replacing the standard username – password authentication scheme with a single camera shot.