Research On Intelligent Face Recognition Access Control System

With the continuous development of computer,information,network and other technologies,while many insecure factors,there is a growing attention to access control security.Now on the market access system has identity with single,security is not high and so on,cannot meet the needs of some high security sites.Biometrics has high stability,uniqueness,and inseparable from the human body,is the more reliable and secure authentication methods.At the same time,with the popularization and the improvement of the performance of smart phones,the mobile-side identity has gradually become a research hotspot.Combining biometrics with mobile phone identity,this paper designs a triple face recognition access control system based on mobile phone,embedded access control and back-end server.The system is suitable for high-security sites.In this paper,based on skin color segmentation and Adaboost two face detection algorithm research and analysis,complementary to both advantages and disadvantages,and be improved to obtain the two fusion algorithms.Firstly,using colour segmentation algorithm to deal with the detected image for the first time processing,get roughly the face regions,then it is used as the detected image of Ada Boost face detection algorithm.By cascade classifier,to achieve high accuracy of face accurate detection.The experimental results show that the fusion algorithm has a higher detection speed,face detection and recognition accuracy is high,low error detection rate;In this paper,aiming the face recognition algorithm based on the classic SIFT feature extraction,try the SIFT feature descriptor is improved,reducing the dimension of the feature descriptor,thereby reducing the complexity of the algorithm.Experiments show that the improved face recognition algorithm can effectively reduce the amount of computation and improve the computing speedThe main innovation of this paper is the design of three verification methods of access control system,the three ways are: face recognition of smart phone,verification of face recognition on embedded and mobile,verification of face recognition on embedded and database.In this article,the face detection and recognition algorithm is embedded into the Android system and embedded system,which is verified by experiments.