Secure Palmprint Verification Using Random Measure and Permutations

Abstract

Compared with the traditional verification schemes, the cancelability and re-issueability of biometric system is the inherent shortages. Therefore, problem of palmprint template security must be considered and solved in a real biometric system. To handle these problems, a hybrid palprint template protection scheme using random measure and permutations is proposed in this paper. Firstly, the Gabor is employed to convolute with preprocessed palmprint image. Then, the obtained Gabor representation is compared with a chaotic random matrix to generate cancelable palmprint template. The random measure can improve the template's discriminablity. At last, the cancelable palmprint template is scrambled to strengthen the security of the template. The publicly available HongKong PolyU Palmprint Database is employed for evaluation. The experimental results show that our proposed scheme can separate the imposter and genuine population very well, that is also to say that, the EER is zero. And the re-usability of the proposed template protection scheme is very large to brute force attack.