Privacy protection method and system for fingerprint authentication

Abstract

The invention provides a privacy protection method and system for fingerprint authentication, and the method comprises the steps: decomposing fingerprint data through employing singular matrix decomposition of low-rank matrix approximation, obtaining a low-rank singular value matrix, and approximately representing a fingerprint matrix; calculating an optimal rank by using an exponential mechanism; then, through truncated singular value matrix decomposition, obtaining a low-rank expression under an optimal rank of the fingerprint matrix; adding Laplace noise to each singular value of the singular value matrix to obtain a disturbance singular value matrix; performing post refinement on the disturbance singular value by using constraint reasoning, and rearranging a singular value sequence to obtain a post-processed singular value matrix; and finally, reconstructing a disturbance fingerprint according to the post-processed singular value matrix. Compared with a similar existing method, the fingerprint privacy protection method has the advantages that fingerprint privacy protection is achieved with minimum information disturbance, irreversibility, non-linkability and accurate identity authentication of fingerprints can be guaranteed more efficiently, and the method has high practicability.

Description

technical field [0001] The invention relates to the cross field of biometric authentication and information security, in particular to a privacy protection method and system for fingerprint authentication. Background technique [0002] Biometric authentication is an extremely convenient and secure way of identity authentication, which uniquely characterizes personal identity through biometrics (such as face, fingerprint, iris, etc.) and behavioral characteristics (such as voice, handwriting, gait, etc.). Among these biometric information, the fingerprint is due to its UUCI characteristics, namely Universality, Uniqueness, Collectability and Invariance, as well as its resistance to theft, forgery, The ready-to-use makes fingerprints known as human ID cards. After being collected and digitized on a large scale, they can be widely used in many practical applications, such as mobile payment, financial management, forensics, and e-government. [0003] However, the UUCI feature o...

AI Technical Summary

Problems solved by technology

[0006] (1) The intensity of added noise is determined by the sensitivity function, and the fingerprint data is relatively more complex and has a higher dimension, so it is more difficult to determine the sensitivity function. Improper addition of noise will cause uncontrollable fingerprint usability;

Images