Research On Palmprint Verification System Based On The Network Security Of Android Moblie

With the rapid development of mobile technology,people can access the network and exchange data by phones at anytime,anywhere.When people enjoy the convenience brought by mobile devices,the occurrence of fraud and identity theft is increasing rapidly.So there is an urgent need for a new security authentication method,which can assist or replace the traditional network security terminal authentication.Biometrics,as an recent developing technology,offers a safe and reliable solution for terminal authentication for the network security.Palmprint recognition technology,as an emerging biometric technology,since its convenience of collecting,the unique and stability of palmprint features and high distinction has attracted people s attention.On this basis,we design a complete palmprint verification system based on the network security of Android mobile.The system has realized three parts,the first part is the acquisition and preprocessing of palmprint image based on Android mobile,the second part is the extraction of palmprint feature and recognition based on the authentication server,and the third part is the communication between Android mobile and the authentication server.The main work in this paper has the following several aspects:1.The acquisition and preprocessing of palmprint image.We can acquire palmprint image through the rear camera of Android mobile.Then we can get the region of interest(ROI,the size of 128×128)of palmprint by a number of steps based on palmprint image,including graying,binarization,morphologica processing,contour extraction,locating and segmentation.Comparing the traditional method and the improved method of valley point positioning,the improved method is adopted in this paper.2.The extraction of palmprint feature and recognition.Considering the size of the received ROI image is 128×128,the down-sampling is used on ROI image to extract effective information and out redundant information in the palmprint image.Then,in order to reduce the impact of light on the recognition result,histogram equalization is applied.After comparing the existing algorithm of the extraction of palmprint feature,PC A and 2DPCA,we select PC A.According to the feature dimension of PCA,K-Neareat Neighbor is adopted as the classification method in this paper.3.The communication between Android mobile and the authentication server.ICE is used to realize the communication,which Android mobile as a client,and PC as a server.There are two parts in this communication.One is sending picture from the client to the server,and the picture is saved by the server.The other is sending recognition result from the server to the client.Considering the different encoding method between Java and C++,the message should be transcoding before being sent.Palmprint image library based on the actual collected is regarded as the experiment test object,which based on the Android platform and VC++ platform.We have achieved good effect in the actual test.