A New Fuzzy Vault Method Using Cubic Spline Interpolation

Abstract

Biometric cryptosystem has emerged as a promising solution in information security. Fuzzy vault is a widely accepted scheme binding biometric data and cryptography effectively. Polynomial projection is often used in the implementation of fuzzy vault. However, the drawback of polynomial projection is that the increase of key length would lead to a higher degree of polynomial, reconstruction of which would cause poor convergence in practice. To resolve the problem, we propose a new fuzzy vault method based on cubic spline interpolation. The piecewise low-degree polynomial of spline interpolation could overcome poor convergence problem and much longer key could be generated. Because the security of the fuzzy vault arises from the number of chaff points in the vault, the same security level could be guaranteed compared to polynomial reconstruction based fuzzy vault. Experimental results based on HA-BJTU palm print database show the feasibility of the proposed method and performances are satisfied.