The Design Of Omni-direction Security Authentication System

Personnel security authentication is an indispensable part of safety certification in some important places, such as chamber meetings, and concealed authentication without contact is the best mode of operation of all safety certification, and omni-direction security authentication meet the needs of safety certification in recent years. The system of omni-direction security authentication designed in this paper is based on facial feature recognition and RFID: the facial feature have the advantages, such as, the non-contact mode, the remote gathering and extracting, which meet the need of security authentication of hidden model; RFID cards each has a 64-bit numbers, which is the only card number, and any two RFID card numbers will not be repeated. Secondly, each RFID card has 16 RFID card encryptions and the complexity of the algorithm determine its safety. And mostly, the information stored in each card is the card owner's facial features, which enhances security authentication system reliability. It is these ensure the omni-direction of the security authentication system.For the security authentication process, first of all, there will be a center for registration as an important part of the security authentication system in the meeting hall, in security checking institution before the meeting. The camera in the registration center capture person's facial image. And then that facial image is changed to eigenface, the feature data characterizing facial image, by the face-extraction algorithm computing. And the eigenface with the natural information of persons registered together is stored into RFID cards. According to the qualification of person registered the system put the corresponding key information into the card that will be given to the person registered. After then when the registration person with the comprehensive safety certification logo enter a security authentication space, the camera hidden take the facial features templates, tracking, location from the face image and transit the image to system to be process with face-extraction algorithm. At the same time RFID readers through the loop antenna hidden read into numbers, and key characteristics of surface model data of the staff. The security authentication system processes them real-timely.In terms of hardware, security authentication system in the hardware has two independent systems: one for the accreditation of the scene authentication system and one for the registration system. Authentication system needs hardware and equipment as follows: capture equipment (LE-8104A Digital Video Capture Calgary CCD color camera JK-2616), radio frequency readers (xiupai send RFID-RF remote reading/writing cards), micro - processing unit (PC with a windows system), display device (sumsung LCD 943N), audio broadcast (Huiwei S3W) and the necessary interface; Registration system needs hardware and equipment as follows: capture equipment, processing units, monitors, RFID card writers.In the term of the algorithm, the principle of security authentication system based on face recognition is as follows: Firstly image is read for face detection processing– separate facial image from the complex background; Feature extraction– get the eigenface from the face image as the identity of authentication; Discrimination– contrast the eigenface and data stored in the RFID card to make sure person, according to the similarity of conditions to meet classifier to identify their status.For the design of the security system in this paper, we have taken the face detection methods based on color to separate the region of facial only image from complex background: Image taken by capture is stored as a BMP image file by computer processing units. After that the BMP file will be read by our application to change into the skin and non-skin pixel input due to the YCrCb model. Then it is processed by face detection algorithm. Finally image output contains face only region.Next we opt for the face recognition algorithm based on PCA: According to the security authentication system described in the paper, the camera capture pictures of each person in the centre of registration. The card distributed to person contains the eigenface of the card owner. The eigenface is computed by the training samples that we build for the system. At the end of registration each person has RFID card with the eigenvector in it. When the card owner get in the place where has authentication system, the camera scene extract the user's facial features, while the radio reader read the card on the user. The authentication system will contrast the two feature to make sure the authentication.In the term of the software application, security authentication system designed in this paper develop at software environment software environment, Visual C + + development platform, used in all C + + language.Finally as summarized, the security authentication system described in this paper is base on face feature and RFID. The facial feature is the basis for the identification, and RFID card with the user identification make the identification reliable, concealed.