Improving fingerprint minutiae matching using local and global structures

This paper focus on the use of local and global structures for fingerprint minutiae matching, we propose a set of modifications and improvements that increase the speed and accuracy. Reducing the number of minutiae pairs and descriptors are the principal amelioration in the local matching step, in the global matching step, we add more minutiae pairs to align the input fingerprint with the template fingerprint for more accuracy. A series of experiments conducted over the four databases of FVC2004 using a fingerprint verification framework, the results show that our improvements makes the novel algorithm more efficient compared to the original algorithm and other algorithms proposed in the literature.