Fingerprints are unique, permanent and universal. The minutiae of a human's fingerprints have sufficient details. We can use these non-trivial details as identification marks to verify fingerprints. The purpose of this paper is to investigate and implement the operation of the minutiae based fingerprint matching system. Minutiae-based fingerprint matching is widely used for fingerprint verification. In this first method, the fingerprint image is enhanced by fast Fourier transform and converted into binary image for further processing. In the second step, the image is thinned and minutiae are extracted. Finally, pairs of minutiae of two fingerprints are matched to obtain the matching score.
Fingerprint recognition has been investigated over a long period of time and has shown the most promising future in real-world application. However, due to the complex distortions between different finger prints in real life, fingerprint recognition remains a difficult problem. Matching two fingerprints may not be successful due to various reasons and also depends on the method that is being used to match. The electronic voting machine (EVM) is a simple electronic device used to record votes in place of ballots and boxes that were used before in conventional voting system. Because biometric identifiers can not be easily mislaid, forged or shared, they are considered more reliable for the recognition of people than traditional tokens or knowledge-based methods. In this paper, the authors are interested in comparing three fingerprint matching algorithms by performing the choice using EVM novel. Based on the result of the choice in terms of coincidence accuracy, the time needed to match, the best algorithm is found for EVM novel. The three matching techniques are direct match, minutiae match, and match based on distance Ratios. We performed the evaluation of the FVC-2000 data sets and the results were observed by performing the choice with the help of these matching techniques and the best matching technique found for EVM novel.