Estimation of singular points in fingerprints images

A new method for singular points estimation on fingerprints is presented. It allows the estimation of such points on noisy images. Almost all algorithms for fingerprint recognition use singular points estimation of the finger pattern. It helps speed up the recognition because the singular points can be used as strong points for fingerprint features matching. Fingerprints classification frequently is also based on localization of such points. Classification is used to reduce the search space size of a fingerprint with the same class, before attempting exact matching.