Design And Implementation Of Face Recognition Identity Authentication System Based On Android Platform

With the development of mobile Internet and artificial intelligence technology,the biometric technology based on personal characteristics has replaced the traditional authentication method.Face recognition is an important part of biometric technology and has unique advantages that other biometric technologies do not have.This paper focuses on the design and implementation of face recognition authentication system based on Android platform.The main work of this article is summarized as follows:1.The original image preprocessing algorithm is introduced and researched.Through the image preprocessing algorithm,the original image's contour is clearer and the contrast is stronger.First of all,the original image is grayscale normalized.In order to reduce the influence of illumination and other factors on the image,the grayscale image is further subjected to histogram equalization processing.After the histogram equalization processing,the image will be More clear and soft,and finally introduced several image filtering solutions to eliminate noise.2.The concept of the Adaboost face detection algorithm is introduced.The feature extraction algorithm is introduced.The feature extraction of the face region can effectively reduce the dimensions of the face image,and the efficiency of the classification and recognition of the next classifier can be mentioned.Accuracy.This article focuses on the feature extraction algorithm based on PCA,and describes in detail the process of PCA feature extraction.3.The concepts and classification algorithms of face recognition are introduced,and the concepts and principles of several common machine learning classifier algorithms are highlighted.Focused on the nearest neighbor classifier algorithm and support vector machine classifier algorithm.By analyzing the use scenario of face identity authentication system and improving the accuracy of face recognition as much as possible,a classifier algorithm based on the combination of nearest neighbor algorithm(KNN)and support vector machine algorithm(SVM)is proposed.After testing on some common data sets,this combined classifier algorithm has higher accuracy and performance advantages in face recognition than using the nearest neighbor algorithm or support vector machine algorithm alone.4.The entire identity authentication system is implemented on the Android platform.After system requirements analysis,the system framework was built according to the characteristics of the Android system.The main application of the Android SDK to achieve the interface and the camera to capture the picture work,using C++language and OpenCV library to achieve the image processing algorithm,and use JNI and Android NDK to combine the two to form a complete system.