Research On Examinee Identity Authentication System Combining RFID With Face Recognition

For the past few years, with the rapid development of modern science and technology, the demands for education and talents are enlarged which increases the difficulty of examination and the pressure of examinees. The phenomenon that is to take exam for someone else is a common occurrence in college examinations, which harms public benefit and influences the authenticity of examinations. How to enhance the accuracy rate of authentication, guarantee the justice and equity and eventually achieve the real intention of examinations is becoming a problem demanding prompt solution which leads to the widespread use of biometric technology, like face recognition. But most of the personal identification technologies are based on single feature which leads to the limitations, like safety loophole and invalidity.This paper describes a new type of exam identity recognition system which combines face recognition and RFID card based on the research of both technology, it is a double recognition system with campus IC card and face recognition, Specific content as follows:(1) The human biological characteristics combined with a face recognition have advantages of portability, invariance, non invasive and as the carrier of data security, RFID card has the advantages of portability, and it is not easy to copy. Based on these, this paper proposes a new method combined of face recognition and examinee identity authentication, it can effectively solve the problems of subjectivity of artificial detection in exam, and the replication of paper certificates, and other problems such as time-consuming. This method reduces the authentication time, ensures the authenticity of the examinee identity and avoids the problem of exam stead.(2) On the basis of in-depth research of face detection methods, we adopt the optimal algorithm, AdaBoost, to the development of face detection module. The experimental results show that AdaBoost has an efficient performance in the novel identification system.(3) On the basis of in-depth research of the advantages and disadvantages of Various face recognition methods, a improve algorithm which combines singu lar value decomposition and modified PCA(ISPCA) is proposed. The experimen tal results show that the modified ISPCA would avoid the problem that light con dition will affect the face recognition process.(4) Based on the above researches, system’s overall framework and processes are proposed, the correctness of the certification process is validate when a prototype system is implemented on PC. As the examinees swiping one-card, the RFID reader will detect the identity information, and then the identification system captures the face image of cardholder. The system will compare the facial feature acquired from face database according to the information in one-card with the captured image. If they matches, then permit; if not, warn.