Shift, rotation and scale invariant optical fingerprint verification system with double random phase encoding

• Published in: Journal of engineering (Stevenage, England) Vol. 2020; no. 13; pp. 476 - 481
• Main Authors: Zhuo, Shengkai, Jiao, Shuming, Zou, Wenbin, Wang, Na, Li, Xia
• Format: Journal Article
• Language: English
• Published: The Institution of Engineering and Technology 01.07.2020
• Subjects: authentication failure; biometric features; biometrics (access control); computer algorithms; cryptography; double random phase encoding optical lens system; DRPE fingerprint optical authentication system; encoding; fingerprint identification; fingerprint image pre‐processing scheme; fingerprint images; image coding; optical systems; scale invariant optical fingerprint verification system; scale variance; slightly distorted fingerprint image; The 3rd Asian Conference on Artificial Intelligence Technology (ACAIT 2019); fingerprint identification; scale variance; fingerprint image pre-processing scheme; scale invariant optical fingerprint verification system; cryptography; encoding; biometrics (access control); double random phase encoding optical lens system; fingerprint images; computer algorithms; DRPE fingerprint optical authentication system; authentication failure; image coding; slightly distorted fingerprint image; optical systems; biometric features
• ISSN: 2051-3305; 2051-3305
• DOI: 10.1049/joe.2019.1142

Biometric features such as fingerprint images can be authenticated by not only computer algorithms but also optical systems. In previous works, double random phase encoding (DRPE) optical lens system is employed for fingerprint verification. However, such a system can yield authentication failure when the key is generated from a slightly distorted fingerprint image. In this work, a fingerprint image pre-processing scheme for DRPE fingerprint optical authentication system is proposed. Simulated experimental results demonstrate the proposed method has significantly better robustness to shift, rotation and scale variance than some of the conventional DRPE-based approaches.