FINGERPRINT IDENTIFICATION USING MINUTIAE CONSTELLATION MATCHING

In this paper we propose a method for fingerprint recognition based on minutiae constellation matching. We use a classic representation for the minutia that consists of a set of parameters including its coordinates and its orientation. Minutiae are grouped in constellations. The constellation formation is achieved by fusion of the entire minutia within a given distance in a unique graph. Fingerprint matching is then considered as a simple problem of graph matching. We first check globally the constellation shape and orientation match before using minutia specification as a second step of verification.