Authors: Munaga V.N.K. Prasad; Jaipal Reddy Anugu
Addresses: Institute for Development and Research in Banking Technology, Hyderabad, India ' Institute for Development and Research in Banking Technology, Hyderabad, India; School of Computer and Information Sciences, University of Hyderabad, Hyderabad, India
Abstract: In this paper we present a multi-algorithmic biometric system comprising the fingerprint biometrics to tackle highly uncertain databases. Two methods were proposed namely, spiral-based method and block-based method to generate cancellable fingerprint templates. Spiral-based method can be used to exploit elements in each contiguous right angled triangle, that are drawn using the invariant distances between reference minutia to every other minutia. In the block-based method, fingerprint image can be tessellated into blocks to get a set of features by considering Euclidean distance from each minutia in reference block to other minutia in blocks which are horizontal, vertical, and diagonal in direction to reference block, and orientation of minutiae. Two biometric methods are fused at score level based on Yager's triangular norm. The experimental evaluations and analysis are performed using the FVC2002 databases. The results show better performance and separation between genuine and imposter distributions.
Keywords: multi-algorithmic biometrics; spiral-based; block-based; tessellation; cancellable template generation; triangular norm; score level fusion; fingerprint models; fingerprints; fingerprint images.
International Journal of Trust Management in Computing and Communications, 2016 Vol.3 No.3, pp.224 - 245
Received: 02 Mar 2016
Accepted: 26 Apr 2016
Published online: 27 Jan 2017 *