Fingerprint template management for higher accuracy in user authentication

This paper proposes effective methods of fingerprint template management for partial fingerprint-based authentication system. Recently, user authentication using fingerprints has become common in mobile devices such as smart phones. Embed type sensors are provided limited space for better product design. Hence a single acquisition can cover only a small portion from the whole fingerprint. To supplement insufficient area, device should store a multiple fingerprint image from a single finger to a template. Since skin deformation by scratch or aging has negative impact on performance, template is required to be updated in a regular basis. And the number of acquisition request to users and a memory size of the device are limited resources which imply efficient template maintenance policy is important concerns. This paper presents a novel template management mechanism using mosaic images to maximize the efficiency of limited template data. We identified the proposed algorithm enhanced the matching accuracy surpassing the general template composing, in our experiments with extensive database of partial fingerprint images of more than 100 people, acquired by a small-sized capacitive sensor.