Fingerprint classification method based on analysis of singularities and geometric framework

According to the former contributions, the authors present a novel fingerprint classification method based on analysis of singularities and geometric framework. First, a robust pseudoridges extraction algorithm on fingerprints is adopted to extract the global geometric shape of fingerprint ridges of pattern area. Then, by use of the detected singularities and with the help of the analysis of the global geometric shape of fingerprint ridges of pattern area, the fingerprint image is classified into different pre-specified classes. This algorithm has been tested on the NJU fingerprint database which contains 2500 images. For the 1000 images in this database, the classification accuracy is 92.2%.