A fingerprint verification algorithm using tessellated invariant moment features

In this paper, an enhanced image-based fingerprint verification algorithm is proposed to improve matching accuracy and processing speed by overcoming the demerits of previous methods over poor-quality images. It reduces multi-spectral noise by enhancing a fingerprint image to accurately and reliably determine a reference point, and then aligns the image according to the position and orientation of reference point to avoid time-consuming alignment. A set of fixed-length moment features, invariant to the affine transform, is extracted from tessellated cells on a region of interest (ROI) centered at the reference point. The similarity between an input and a template in a database is evaluated by eigenvalue-weighted cosine (EWC) distance. Experimental results show that the proposed method has better performance in accuracy and speed comparing with other renowned methods.