Approximate fingerprint matching using kd-tree

Fast and robust fingerprint matching is a challenging task today in fingerprint-based biometric systems. A fingerprint matching algorithm compares two given fingerprints and returns either a degree of similarity or a binary decision. Minutiae-based fingerprint matching is the most well-known and widely used method. This paper reveals a new technique of fingerprint matching, using an efficient data structure, combining the minutiae representation with the individual usefulness of each minutia, to make the matching more powerful. Experimental results exhibit the strength of this method.