Postprocessing algorithm for security features extraction

One of the issues of fingerprint recognition is finding minutiae or ridges, which involves deciding whether the pixel evaluated is a valid minutiae (ridge ending or ridge bifurcation) or not. Usually, the minutiae are detected in the thinned image which contains a large number of false minutiae and noises. Its may highly decrease the matching performance of the system. This paper proposes an algorithm which is used to eliminate spurious minutiae and non-component of fingerprint features.