A Convolutional Neural Network based Fingerprint and Face Biometric Multi-modal for Educational Authentication System

  • Mr. R. Sivakumar, R. Sivakumar

Abstract

Unimodal biometric systems (UBS) have little efficiency in classifying people, mainly because of their susceptibility to changes in individual biometric characteristics and attacks at presentations. The multimodal biometric systems used to identify a person have attracted the attention of researchers because of their advantages, such as higher recognition efficiency and higher security compared to UBS. To crack a biometric multimodal system, an attacker will have to crack more than one single-mode BS. In multimodal biometric systems: having multiple functions means that a multimodal system is becoming more reliable. Multimodal BS increases security and ensures the confidentiality of user data. This article presents a fingerprint and biometric multimodal face model for the educational authentication system based on the convolutional neural network (CNN). The proposed module has been tested with online data and student data in real time. Finally, performance parameters are calculated in terms of recall (R) and F measure (FM), accuracy (ACC), precision (P). The proposed algorithm has reached 99.28% accuracy for online data. For these students in real time, the proposed system has archived 97.54% accuracy. And also our methods provides a high authentication in educational sector. Which is better than with compared to various existing methods.

Published
2020-02-20
How to Cite
R. Sivakumar, M. R. S. (2020). A Convolutional Neural Network based Fingerprint and Face Biometric Multi-modal for Educational Authentication System. International Journal of Advanced Science and Technology, 29(3), 3486 - 3497. Retrieved from http://sersc.org/journals/index.php/IJAST/article/view/4792
Section
Articles