FINGERPRINT SUB-CLASSIFICATION AND SINGULAR POINT DETECTION

This paper presents a novel approach to fingerprint singular point det ection. Singular points (cores and deltas) are used for fingerprint cl assification, sub-classification and registration. This method exploit s the stability of the directional field pattern extracted from singul ar point regions at different resolution levels. The procedure is inva riant to translations, scaling and small rotations. A fingerprint sub- classification procedure was built based on the proposed singular poin t detection method. Two kinds of tests were conducted on a subset cons isting of 955 NIST-14 fingerprint images. First, automatic and forensi c expert sub-classifications were compared. Second, the consistency of the proposed method was measured comparing automatic sub-classificati on for two different rolls of the same fingerprint.