17 May 2011 Adding localization information in a fingerprint binary feature vector representation
Author Affiliations +
Abstract
At BTAS'10, a new framework to transform a fingerprint minutiae template into a binary feature vector of fixed length is described. A fingerprint is characterized by its similarity with a fixed number set of representative local minutiae vicinities. This approach by representative leads to a fixed length binary representation, and, as the approach is local, it enables to deal with local distortions that may occur between two acquisitions. We extend this construction to incorporate additional information in the binary vector, in particular on localization of the vicinities. We explore the use of position and orientation information. The performance improvement is promising for utilization into fast identification algorithms or into privacy protection algorithms.
© (2011) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Julien Bringer, Vincent Despiegel, Mélanie Favre, "Adding localization information in a fingerprint binary feature vector representation", Proc. SPIE 8029, Sensing Technologies for Global Health, Military Medicine, Disaster Response, and Environmental Monitoring; and Biometric Technology for Human Identification VIII, 80291O (17 May 2011); doi: 10.1117/12.884787; https://doi.org/10.1117/12.884787
PROCEEDINGS
10 PAGES


SHARE
Back to Top