The Research Of Multi-biometric Encryption Method Based On Fuzzy Vault

With the development of the whole social network, network security has become a key issue during network development. Biometric identification technology as the fastest growing and the most potential a kind of information security technology, its application is to be more and more widely. With the application of biometric technology becoming more in-depth, itself the inherent defects and deficiencies have been exposed. The main defect is the privacy of the biometric involved and the security issues. Therefore, a biometric encryption technology which combine biometrics and password technology emerged as the times require, and it became an important development trend in the future information security area. At the same time through repeated certification, multi-biometric fusion encryption method is one of the hotspots of biometric encryption technology.In this thesis, we are starting from two single-mode authentication technology of fingerprint recognition and face recognition. And we make a more detailed discussion and result for multi-biometric fusion encryption method. Our main work is listed as follows:1. Fingerprint recognition correlation algorithm is researched. Firstly, we can get thinned fingerprint image, by preprocessing the input fingerprint image. And then extract the minutiae. The quality of minutiae extraction will directly affect the process of later image matching.2. Face Recognition based on eigenface technique is researched. We analyses and discuss the several methods of face recognition, comprehensive comparison the advantages and disadvantages of these methods. And then we propose an improved PCA face recognition method, which can extract face feature simply, fast.3. Fuzzy vault algorithm is researched. First we introduce biometric encryption technology classical method. Then we discuss and research the traditional Fuzzy Vault algorithm in detail. And we analyze the corresponding loopholes and vulnerabilities in terms of security and performance. Finally we propose an improved algorithm, and further enhance the security and flexibility of the algorithm.4. Multi-biometric encryption model is researched. We first research and analysis the multi-biological information fusion. According to face feature large intra-class variations we present a hybrid face template protection method. Then this paper is the direct use of biometric encryption algorithms, namely Fuzzy Vault algorithm to research multi-biometric encryption. At the same time, we propose an AND model for multi-biometric encryption model in a decision level fusion.In general, we propose an improved Fuzzy Vault algorithm to improve biometric security and flexibility. And at the same time, this algorithm is applied to multi-biometric encryption of the fusion of fingerprint and face. Then we analyze the security of the whole algorithm and get a good recognition effect.