Shanthini K. Sreenivasan and Deepthi P. Pattathil
Biometric encryption, Finger print, Ridge feature map, Gabor filtering, Authentication
Biometrics-based authentication systems that use physiological and/or behavioural traits (e.g., fingerprint, face, and signature) are good alternatives to traditional methods. These systems are more user-friendly and more reliable since biometric data cannot be lost, forgotten, or guessed. When cryptography is combined with biometrics, the digital signature created by a user is linked with him/her more accurately. Two methods are proposed here for reliable fingerprint authentication. In both methods, helper data is created by linking a key with the user’s fingerprint, which can be saved in a secure token device such as a smart card and this do not leak any information about the key. The key can be reproduced upon submission of both biometric and the token. It is quite difficult for an attacker to procure both and derive the key. Gabor filtering is applied to represent a fingerprint image as a compact sized data called ridge feature map. The major benefit of the ridge feature map is that it does not require any kind of fingerprint alignment. Then error correction is applied to accommodate the variability in biometric data using look up table and using error control coding. The feasibility of the technique in practical usage is being verified by the experimental results and the security analysis in DB1 (FVC 2000) dataset.
Important Links:
Go Back
