Vol.63, No.1, 2020, pp.17-30, doi:10.32604/cmc.2020.09044
Left or Right Hand Classification from Fingerprint Images Using a Deep Neural Network
  • Junseob Kim1, Beanbonyka Rim1, Nak-Jun Sung1, Min Hong2, *
1 Department of Computer Science, Soonchunhyang University, 31538, Asan, Korea.
2 Department of Computer Software Engineering, Soonchunhyang University, 31538, Asan, Korea.
* Corresponding Author: Min Hong. Email: mhong@sch.ac.kr.
Received 07 November 2019; Accepted 18 December 2019; Issue published 30 March 2020
Fingerprint security technology has attracted a great deal of attention in recent years because of its unique biometric information that does not change over an individual’s lifetime and is a highly reliable and secure way to identify a certain individuals. AFIS (Automated Fingerprint Identification System) is a system used by Korean police for identifying a specific person by fingerprint. The AFIS system, however, only selects a list of possible candidates through fingerprints, the exact individual must be found by fingerprint experts. In this paper, we designed a deep learning system using deep convolution network to categorize fingerprints as coming from either the left or right hand. In this paper, we applied the Classic CNN (Convolutional Neural Network), AlexNet, Resnet50 (Residual Network), VGG-16, and YOLO (You Only Look Once) networks to this problem, these are deep learning architectures that have been widely used in image analysis research. We used total 9,080 fingerprint images for training and 1,000 fingerprint to test the performance of the proposed model. As a result of our tests, we found the ResNet50 network performed the best at determining if an input fingerprint image came from the left or right hand with an accuracy of 96.80%.
Deep Learning, convolution neural network, fingerprint classification.
Cite This Article
Kim, J., Rim, B., Sung, N., Hong, M. (2020). Left or Right Hand Classification from Fingerprint Images Using a Deep Neural Network. CMC-Computers, Materials & Continua, 63(1), 17–30.
This work is licensed under a Creative Commons Attribution 4.0 International License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.