Fingerprint based person identification & verification for commercial applications

This paper presents the development of a simple automated fingerprint identification & verification system. The technique conferred in this paper is based on the extraction of feature vectors from the 2D FFT of the binarized image of a fingerprint using a wedge-ring extraction process. The technique used can accommodate a small degree of shift and rotation (within 15") in the placement of a fingerprint. The mechanism detailed in this paper also possesses a good recognition speed and therefore it can be used in conjunction with a fingerprint scanner for on-line fingerprint recognition. Simulation results obtained indicate low false acceptance (1%) and false rejection (5.5%) ratios. The proposed method is found to be reliable for systems with a small to medium sized set of fingerprint data.