Design And Implementation Of Face Recognition System For Entry And Exit Frontier Inspection

Compared with other biometric features,the face feature recognition method is more friendly,more subtle,more direct,more convenient,and more difficult to be deceived.In the case of the object is not aware of the object,will not lead to identify the object of resentment,and more easily accepted.In frontier inspection,through the layout of the surveillance video,the entry-exit personnel are not aware of the situation,you can recognize the facial features,in order to improve the security and immigration,immigration management efficiency.First of all,the depth learning algorithm is used to solve the problem of human face recognition.First,through the positive conversion of face recognition,to obtain a variety of abstract identification of input samples,on this basis,build and construct the reverse model,each layer of input samples,the final adjustment between each layer through representation and target,obtained with a deep learning model for the.And the deep learning algorithms,data integration,data cube construction regulation RBM(RBM)and feedback tuning steps,by comparing the experimental results with the PCA to face recognition algorithm can be seen,the deep learning recognition algorithm based on high accuracy not the good recognition effect,image identification of fluctuations is smaller.Secondly,on the whole frontier inspection Kennedy face recognition system construction research,which mainly includes video surveillance face recognition on the entry and exit frontier inspection station is mainly responsible for the collection of video image data acquisition and video compression encoding part to monitor the acquisition terminal data acquisition terminal,video data of video data storage access management the end of the study.And focus on the use of OpenCV technology in the face positioning module,and the use of deep learning algorithm to achieve the face recognition module.Through the immigration checkpoint face recognition system test results show that the accuracy of face recognition system is studied in this paper in more than 95%,and the performance of the system is also able to meet the needs of the performance of face recognition immigration inspection.