Extended Feature Extraction Technique From Fingerprint

Abstract

Automated Fingerprint Identification Systems (AFIS) currently rely only on Level 1 and Level 2 features. But these features are not much helpful for forensic experts as the experiment deals with partial to full print matching of latent fingerprint. Forensic experts takes the advantage of extended feature proposed by "Committee to Define an Extended Fingerprint Feature Set" (CDEFFS). This paper presents extraction technique of two extended features, dots and incipient ridges by tracing valleys. We have found starting points on the valley by analyzing the frequencies present in the fingerprint. Valley are traced from the starting point using first marching method ( FMM ). Then an intensity checking method is used for finding these features. Extensive simulation is carried out in MATLAB environment to show the superiority of the proposed feature extraction technique over state of art. Accuracy of the proposed feature extraction scheme also has been shown using special database 30 and IIIT Delhi database.