SIFT-based algorithm for fingerprint authentication on smartphone

Abstract

Fingerprint authentication with small area sensor "touch sensor" becomes the most promising technology for network user authentication on mobile devices such as smart phones. In this situation, the size of touch sensors becomes so small that the conventional minutia method should be replaced with a new approach. We consider Scale Invariant Feature Transform (SIFT) approach for fingerprint authentication with a touch sensor on smart phones. In this paper, we focus on template expansion on registration in order to accept any small part of the query finger for verification.