Fingerprint recognition system to verify the identity of a person using an online database

Abstract

Our B.Tech project emphasize on the current techniques for the fingerprint recognition. Human fingerprint exhibit some certain details marked on it. We categorized it as minutiae, which can be used as a unique identity of a person if we recognize it in a well manner. The main aim of this project is to design a complete system and an indigenous design model for fingerprint verification from an online database using minutiae matching technique. So, in order to have a good quality minutiae extraction the fingerprint image is first pre-processed by image enhancement which includes histogram equalization, Fast Fourier Transform and binerization and then segmentation is done to get the effective area of the fingerprint followed by minutiae extraction which includes ridge thinning and minutiae marking and then we have a post-processing operation which includes removal of H-breaks, isolated points and false minutiae. Now, we go for a final treatment which is ‘minutiae matching’, in minutiae matching we match the post-processed fingerprint image with the online database.
For all these operations, we develop an alignment based matching algorithm which is for minutiae matching. This algorithm has a specialty that it itself finds the correspondences between input minutiae and the stored template minutiae pattern and there is no resorting to exhaustive search. We can then evaluate the performance of the system on a database by taking fingerprints of different people.