Research On Multi-Biometric Encryption Algorithm Based On The Feature Level Fusion Of Fingerprint And Finger Vein

Biometrics is one of the most widely used technologies in the world for identification schemes.The applications of biometrics such as fingerprint and finger vein can often be seen in electronic devices in different life scenarios.However,this kind of recognition technology that relies on a single biometrics has some limitations.For example,some users may lack a certain biometrics,which makes the performance of the single biometrics recognition technology far better than that of the multi-biometrics recognition technology in terms of both recognition performance and security performance.But,whether single or multiple biometrics,biometric identification is a biometric template the possibility of loss,especially more biometric identification technology,the internal database to store a large amount of biometric template in the relevant to the user,in the event of lost feature template,lead to privacy problems will become more serious.In order to protect the biometric template,enhance the biometric template of biometric identification system security and reliability of the identity authentication,fuzzy vault of biometric encryption system is commonly used in solution,and make the improvement will be applied to many biometric encryption,is proposed based on the fingerprint and refers to the vein feature level fusion of multiple biometric encryption algorithm,the following is the detail content:(1)this paper focuses on the study of various biometric encryption schemes,and explains the process of each encryption scheme,especially the algorithm process of the fuzzy vault scheme.Then,the fusion levels of multi-biometrics fusion technology are studied,the fusion technologies used in multi-biometrics recognition system are summarized,and each fusion technology is analyzed and compared in terms of information,difficulty of implementation and security of biometrics templates.Finally,each part of the fingerprint feature point extraction algorithm and the fingerprint vein feature point extraction algorithm is realized,including the fingerprint image preprocessing and fingerprint detail point extraction,the fingerprint vein image preprocessing and finger vein detail point extraction.(2)aiming at the security problems existing in the existing biometric encryption system,a new fast fuzzy vault algorithm based on the fusion of fingerprint and digital vein biometric was proposed.The feasibility,reliability and safety of the fusion scheme are analyzed based on the experimental results.In the algorithm,the fingerprint feature binary matrix is firstly extracted,and the feature template is retracted by combining with the random matrix generated by the system in an xor way.The binary matrix transformation is carried out by using the random sequence to avoid the relevant attack in the fuzzy vault.Then based on relative difference refers to the vein detail points to generate the key refers to the vein,and by building a fast fuzzy vault fused two biological characteristics,also in fuzzy vault algorithm by increasing the distinguishing characteristics of types,reduce the declassification of systemtime,use refers to the way of vein key ratio,improved the security of the CRC system when collisions can occur.