Design and Implementation of Gabor Filter and SVM based Authentication system using Machine Learning

The most vital requirement in today's world is to overcome the different types of attacks. Human behavioral and physiological features in biometrics have the largest scope as a solution for security issues. However, the existing biometric systems such as faces, iris, palm, voice or fingerprints are highly complex in terms of time or space or both, and thus are not suitable in high security. So the design and implementation of finger-vein authentication method is proposed in this paper. This system is implemented using a combination of image processing and machine learning algorithm. Lacunae, fractal dimension and gabor filter are the algorithms used for feature extraction and the classification of the extracted feature is done using the Support Vector Machine. The accuracy of classification algorithm for One-Versus-One and One-Versus-All is 98.75 % and 97.92 % and the execution time is 0.168 Seconds and 0.187 Seconds respectively. At the end the comparative analysis between different classification algorithm and previous research work related to Finger Vein Authentication System using Machine learning is provided.