Silhouette-based hand recognition on mobile devices

Hand Biometric Recognition not only gathers a good performance in identifying individuals but also it is known to be a non-invasive biometric technique. Furthermore, there exist new trends towards mobile implementation developments, focusing on embedding current biometric systems in mobile devices. This paper aims to implement hand biometric recognition with mobile devices. This approach involves a completely contact-less acquisition process, since the mobile devices only requires a low-quality camera. Moreover, a shape-based template provides a unique pattern for identification and verification, far different from previous implementations focused on diverse geometry measurements for efficient recognition. Results highlight the fact that Hand Biometric Recognition can be implemented in mobile devices with a considerable accuracy in identification/verification rates, being this implementation one of the first biometric identification systems carried out in mobile devices, considering moreover the fact of being without any contact and with low quality images. Posterior applications could take advantage of this implementation, since this model allows to extract a cryptographic feature from the hand template, or being combined with gesture recognition for aliveness detection in mobile platforms.