Research And Implementation Of Security Management System Based On Face Recognition

In recent years,with the prosperity and development of social economy and the progress of science and technology,more attention paid to the person's identification field.Compare to the traditional manual way,the face recognition has higher security,is currently a hot research direction of artificial intelligence.Based on the Open Source Computer Vision Library and face recognition,this paper aims at develop a safety management system,combined with the traditional way of ID card verification,effective management and control of the import and export of prison The system mainly includes access control terminal,remote server management terminal and handheld Android device management terminal.The access control terminal collects face image and ID information,and sends these to the server through the high-speed network to face detecte and recognite,and returns the result to the gate control terminal to perform the response.The handheld Android device manager can view real-time or historical entry and exit situations at any time and handle them when there is an exception.In this system,face recognition is done in the order of face preprocessing,face detection and face recognition(training).Face preprocessing mainly includes the steps of image grayscale,image filtering,histogram equalization,and so on.Face detection using AdaBoost based face detection algorithm.The main face recognition algorithm is studied: the LBPHFace face recognition algorithm based on LBP.This algorithm is used for face recognition.This paper completes the design and implementation of a security management system based on face recognition.After real-time testing and accuracy testing,the system runs well,and the real-time and accuracy rates meet the requirements.