Research On 3D Face Recognition System Based On Android System

With the rapid development of artificial intelligence and Internet,network payment,identity authentication and other technologies are gradually popularized,and the problem of information security is becoming increasingly serious.Because of the advantages of non-contact,naturalness and convenience,face recognition has become the research focus in the field of identity recognition.However,two-dimensional face recognition has the disadvantage of being easily faked by photos or animated pictures,and three-dimensional face depth data contains face depth.Information can effectively prevent the occurrence of fraud,more secure and convenient,and can well protect the security of user information.This article first combines the LBP algorithm with the SVM algorithm and applies it to the field of 3D face recognition.The recognition rate of the test on the Texas 3DFRD database can reach 96.83%,the recognition time is 1.88 seconds,In order to improve the recognition time of face recognition,after combining the LBP+SVM algorithm,an improved MB-LBP+SVM three-dimensional face recognition method is proposed.The simulation experiment was conducted on the Texas 3DFRD database,the recognition rate was 96.88%,and the recognition time was 0.025 s.The recognition rate in the self-made depth library is 96.36%,and the recognition time is 0.02 s.It can be seen from the experimental results that the simulation results on the self-made face database and the Texas 3DFRD database are consistent.The hardware platform is built with D435 i camera,RK3399 development board and LVDS signal high-definition display.The Android system is set up in the hardware platform,and the code of the simulation experiment research algorithm in this paper is transplanted to this platform.On this hardware platform,two APPs were made using Qt Creator tool and Android Studio tool,they are responsible for completing the image acquisition and image processing functions,and finally use the Texas 3DFRD database and self-made face database to perform functional tests and performance tests on the transplanted algorithms.The recognition rate on the Android system is not much different from the simulation research,and it still has a high recognition rate.