Embedded Mobile Payment System Based On Face Recognition

With the rapid development of modern communication technology,4G network era has arrived.The improved quality of communication provides a strong support for the mobile phone business platform.Mobile phone is not only a simple communication tool,but also a set of portable platform for communication,multimedia,games,shopping and payment.In the process of mobile payment,the accuracy of identification,the security of the transaction information and the simplicity of operation are the must.As the inherent nature of human beings,the face is constant invariance.At the same time,the human face is the face database In this paper,the AdaBoost is used to determine the location and size of the facial information,and the subspace projection algorithmThe system concludes two aspects,the design of the Server terminal on Windows platform and the design of the Client terminal on Android platform.Server terminal is used for face recognition processing and return identification results.Client terminal is a payment system for collecting and uploading the face information and processing trade information.The main contents are as follows:(1)Through the understanding and learning of the Android platform,this paper expounds the development history of different versions of the Android platform and the system architecture of the Android platform(2)The development progress of face recognition technology and the wide application in different environment.(3)Background face recognition based on Windows servers can be divided into human face information location,discriminant feature extraction and database retrieval.In this system,AdaBoost is used to locate the area and size of human facial information,and the subspace projection algorithm based on principal component analysis is used to identify human identity through facial information.(4)Based on the above knowledge of face recognition and the research on Android platform,a complete payment system for Android platform is designed and implemented.(5)In the weighted feature scheme,different weights are given to different areas of facial information by using the method of weighted feature range.Among them,the weights of feature regions which have obvious influence on recognition effect are larger,while the weights of other regions are smaller,which can effectively improve the accuracy of face recognition.Finally,the system realized the function of face information collected by the client terminal,and face recognition by the server terminal.Through the debug results,our system can obtain the ideal identification accuracy as well as the stability of the system.As the process of image acquisition in natural conditions is easily affected,the further study and improvement is neededApplication Scenario:The original way of attendance and payment is manual card swiping mode.In this mode,people from outside will be mixed in,work cards will be lent to each other,and work cards will be brushed indiscriminately.Identity is identified by face recognition to determine the identity of personnel and to eliminate the chaotic brushing in factories.Using weighted feature scheme to improve the accuracy of recognition can reduce the cost of equipment.In order to increase reliability and security,face recognition generally includes biopsy detection.