Fingerprint Recognition using Scale Invarient Feature Transform (SIFT)

The main object of this project to design a fingerprint recognition system which can perform the task of fingerprint verification with greater accuracy. One of the problems of fingerprint authentication system that it perform poorly in case of poor quality fingerprint image. Extraction of minutia from the poor quality fingerprint image is difficult and not all the minutia is extracted. Different number of minutia fingerprint image creates the problem of accurate matching. Moreover most of the fingerprint recognition system performs poorly in case of rotated version of the query fingerprint image. In this project a Scale Invariant Feature Transform (SIFT) based fingerprint recognition system has been designed to overcome above-mentioned problem. In this system, first, minutia’s or called featured points are extracted from the fingerprint images. These minutia points are then used to match the identity of the fingerprint image. The SIFT algorithm is basically an approach for extracting distinctive invariant features from images. Since SIFT is rotation invariants therefore it gives better result in case of rotated fingerprint image. Simulation results that this system is able to verify the fingerprint with good accuracy.