| With the rapid development of emerging internet technology such as internet of things,cloud computing and big data,information security is becoming an urgent issue in our information driver society for nowadays.The security of traditional cryptography completely rely on the security of cryptographic key,and the cryptographic key has no inevitable relation with legal user.Hence,the cryptosystem will lack the assurance needed about the user’s identity when cryptographic key is lost or stolen.Biometrics can compensate for the shortcomings of cryptographic key in traditional cryptography.However,due to the uniqueness and permanent association of the biometric system,the compromised of biometric templates cannot be used to other biometrics system.Biometrics encryption is an organic combination of biometrics and cryptography,which effectively solves the problem of data security while protecting the cryptographic key.At present,the core problem of biometrics cryptography is how to overcome the contradiction between the fuzziness of biometric feature information and the accuracy of cryptography.In recent years,more and more scholars at home and abroad have been devoted to the research in biometrics cryptography,and fingerprint which is the biological characteristics of oldest application and highest comprehensive performance.Therefore,our research will be developed on fingerprint based key generation technology.This thesis mainly studies on the field of fingerprint based key generation technology,and the contributions are as follows:1.In order to effectively unify the fuzziness of biometric feature and the accuracy of cryptography,a key generation scheme from fingerprint using statistical quantization mechanism and two-level error correcting technique is proposed.In this thesis,the fingerprint images are processed by using statistical quantization mechanism to reduce intra-variation of the fingerprint samples,and extract the reliable genuine fingerprint features.Thus,compensate the fingerprint feature based on interval mapping to generate a stable fingerprint key.Finally,an auxiliary data that have been used to regenerate the fingerprint key is obtained by the processing of two-level error correcting technique.We implement this algorithm in programming,and analyze the feasibility and security with experimental results.Experimental results demonstrate that this algorithm can generate a stable cryptographic key and ensure the security of the encryption key.2.This thesis introduces the process of fingerprint image preprocessing and fingerprint feature extraction in detail,and a key generation algorithm based on relative difference between fingerprint minutiae is proposed.A new feature template is constructed according to the relative distance,relative orientation field and type difference of the fingerprint minutiae.Then,mapping the new feature template into a predefined three-dimensional array,and quantify the rectangular cube which is in the three-dimensional array to generate the fingerprint key.Finally,the helper data that have been used to regenerate key is obtained by the processing of fingerprint key using Reed-Solomon error correcting encoding technique.We implement this algorithm in programming,and analyze the feasibility and security with experimental results.Experimental results demonstrate that this algorithm can generate a longer fingerprint key,with a lower error rejection rate(FRR). |
