A biometric-based system for online security

Abstract

Biometric security systems continue to play an important role in multifactor authentication and are found to be more beneficial as a form of authentication method. Since password authentication alone has found to be lacking, there is a need to include another layer of authentication to improve online security. Fingerprint recognition is selected as a subsequent authentication after password authentication as it is considerably more user-friendly, cost-effective and requires less storage space as compared to other forms of biometric systems. For the project implementation, different databases were used to compare and analyze the results that were achieved. The author adopted a minutiae-based fingerprint biometric system that did feature extraction using Crossing Number based method in MATLAB. Feature matching was done by computing the Euclidean distance between features of 2 fingerprint images to determine if there is a match. MATLAB, in conjunction with MSSQL and Visual Studio, allowed the execution of a multifactor authentication system for online security. However, limitations can still be found in the current project implementation. Hence, it is vital that continuous improvements are made for future developments of fingerprint recognition algorithms.