Research On Fingerprint Encryption Algorithm And Application

In the 21 st century,information security and personal identification have been widely concerned.Cryptography and fingerprint recognition are the key technologies to solve these problems,which have attracted the attention of many researchers.Fingerprint encryption technology is a new technology which is the combination of cryptography and fingerprint recognition technology,and it has brought new solutions to solve the key and fingerprint feature template security issues.In recent years,fingerprint encryption technology has been rapid development,in many fingerprint encryption technology framework,the combination of fingerprint and key has been commonly used,the method of the combination of fingerprint and key fingerprint can be divided into three types,key release,key binding and key generation.Among them,key binding becomes the most popular fingerprint encryption mode because of its high security and its extensive research.In the key binding mode,Fuzzy Vault has a good balance between the fuzziness of fingerprint recognition and the accuracy of cryptography.Now,Fuzzy Vault is the most widely used fingerprint encryption algorithm.The use of fingerprint recognition in smart phone become more and more popular,so the research of the fingerprint encryption in smart phone is urgent.The Fuzzy Vault algorithm for traditional fingerprint has been not suitable for small and medium area fingerprints because of the smaller fingerprint image and the less fingerprint feature.This paper focus on the problem of small and medium area fingerprint encryption,we use an improved Fuzzy Vault algorithm to get the small and medium area fingerprint encryption work.The main work of this paper are as fellow:(1)This paper introduces the development background and research status of fingerprint recognition technology and fingerprint encryption technology,and summarizes the basic framework and technical foundation of the two technologies.(2)In this paper,a method of orientation field extraction based on frequency correction is proposed.The algorithm uses the frequency correction method to correct the wrong orientation field,and the algorithm effectively reduces the number of the pseudo minutiae.(3)In this paper,a method named Symbiotic Compensation Fuzzy Vault(SCFV)is proposed.To solve the problem of insufficient minutiae,this algorithm uses the gained minutiae to generate the symbiosis points,which compensates the set of fine points for encryption and satisfies the basic conditions of the Fuzzy Vault algorithm.(4)This paper realizes a fingerprint encryption software based on small and medium area fingerprint encryption technology to implements the algorithm which is presented above.The algorithms are tested on the self-built database XDFinger,and the results show that for the fingerprint images which have a lot of scars,the average true minutiae ratio extracted by our field extraction algorithm is 59.5%,which is better than 19.6% and 26% of the two traditional algorithms.For the small and medium area fingerprints,the SCFV proposed in this paper reduces the FTCR(Failure-To-Capture Rate)from 38% to 4% under the same security as the traditional Fuzzy Vault algorithm,and increased the GAR(Genuine Accept Rate)from 45.2% to 65.4% at the same time.In summary,the performance of the proposed fingerprint encryption method has been improved significantly in the small and medium area fingerprints,and to a certain extent,the security application of mobile payment technology based on fingerprint recognition is promoted.