The Research And Implementation Of Multi-biometric Cryptographic Technology

The multi-biometric recognition technology improves the recognition rate and the reliability of the identification systems,but in the meantime more biometric templates have to be managed carefully.Managing and Protecting these templates effectively to solve users' needs for safety collection,safety use and safety storage is the key to massively generalize multi-biometric identification system.The uni-biometric cryptographic technology offers a reliable solution for key security and template security issues in uni-biometric recognition technology,and it also provides a theoretical and practical basis for multi-biometric cryptographic technology.In general,the multi-biometric cryptographic technology is still in a starting stage,and it is a challenging field which is worth researching.This thesis mainly studies on the multi-biometric cryptographic technology field,and the main contributions of this thesis are as follows:(1)This thesis introduces the background and the research status of the biometric recognition technology and the biometric cryptographic technology in detail,summarizes the system classification,fusion hierarchy of the multi-biometric recognition technology and fusion level of the multi-biometric cryptographic technology.Then the locking and unlocking interfaces in the fuzzy vault scheme are implemented for our future work.(2)This thesis describes the fingerprint recognition technology in detail,especially the implementation of fingerprint image preprocessing,feature extraction and quantization.Two multi-sample feature-level fusion fuzzy vault schemes using features fused by two feature-level fusion methods are proposed.One is fusing the minutiae in first fingerprint with the nearest minutiae in second fingerprint.The other is fusing the minutiae in first fingerprint with the orientation in the second fingerprint.Inspired by the theory of the Diffie-Hellman key-exchange protocol,the two fusion methods proposed work on a finite multiplicative group and the security of these methods depends on the discrete logarithm problem.We combine the two fusion methods and the fuzzy vault scheme together and implement it in programming using the interface provided above.We analyze the feasibility,reliability and the security of these proposed schemes with the experimental result.It shows that the two feature-level fusion methods are feasible and reliable,and the two multi-sample feature-level fusion fuzzy vault schemes can enhance the security of templates protection significantly,but the security of key protection still has room to improve.(3)This thesis describes the finger vein recognition technology in detail,especially the implementation of finger vein image preprocessing,feature extraction and quantization.A novel multi-modal key-level fusion fuzzy vault scheme using fingerprint and finger vein are proposed.We implement this scheme in programming using the fuzzy vault interface provided above and then analyze the feasibility,reliability and the security of the proposed scheme.The experimental result shows that this multi-modal key-level fusion fuzzy vault scheme is feasible and reliable,and it can enhance the security of key protection significantly and the security of template protection partly.