A New Hand Based Biometric Modality & An Automated Authentication System

  • Nirmal Pandey Delhi Technological University, Delhi
  • Om Prakash Verma
  • Amioy Kumar

Abstract

With increased adoption of smartphones, security has become important like never before. Smartphones store confidential information and carry out sensitive financial transactions. Biometric sensors such as fingerprint scanners are built in to smartphones to cater to security concerns. However, due to limited size of smartphone, miniaturised sensors are used to capture the biometric data from the user. Other hand based biometric modalities like hand veins and finger veins need specialised thermal/IR sensors which add to the overall cost of the system. In this paper, we introduce a new hand based biometric modality called Fistprint.  Fistprints can be captured using digital camera available in any smartphone. In this work, our contributions are: i) we propose a new non-touch and non-invasive hand based biometric modality called fistprint. Fistprint contains many distinctive elements such as fist shape, fist size, fingers shape and size, knuckles, finger nails, palm crease/wrinkle lines etc. ii) Prepare fistprint DB for the first time. We collected fistprint information of twenty individuals - both males and females aged from 23 years to 45 years of age. Four images of each hand fist (total 160 images) were taken for this purpose. iii) Propose Fistprint Automatic Authentication SysTem (FAAST). iv) Implement FAAST system on Samsung Galaxy smartphone running Android and server side on a windows machine and validate the effectiveness of the proposed modality.

The experimental results show the effectiveness of fistprint as a biometric with GAR of 97.5 % at 1.0% FAR.

Published
2018-12-06
How to Cite
Pandey, N., Verma, O., & Kumar, A. (2018). A New Hand Based Biometric Modality & An Automated Authentication System. Journal of Information Sciences and Computing Technologies, 8(1), 750-768. Retrieved from http://scitecresearch.com/journals/index.php/jisct/article/view/1648
Section
Articles