Research And Implementation Of Face Recognition System Based On Android Platform

Face recognition is a kind of reliable and effective biometrics technology,which is to classify and identify people according to some features of their faces,so as to recognize different people.Face recognition has such good features as non-intrusion,convenience,friendliness,non-contact and scalability,etc.It has a large development space and can be applied in many fields.Therefore,face recognition technology develops rapidly.Android is an open source operating system for mobile devices,led by Google and developed by a large number of users.Therefore,the combination of face recognition technology and Android platform will have a broad application prospect.This paper mainly studies the application of face recognition in Android platform,aiming to improve the anti-interference ability of face recognition under complex lighting conditions.This paper mainly includes the following three aspects:(1)Studied the existing image preprocessing,studied the Retinex theory according to the image characteristics,and applied Retinex characteristics to the image preprocessing,combined AdaBoost algorithm with Retinex,improved the traditional AdaBoost algorithm,trained the classifier with the extracted light invariant features,and achieved better face detection.(2)Considering the influence of illumination,this paper chooses the Local Binary Pattern(Local Binary Pattern,LBP)for face recognition,the algorithm of complex environment lighting has a certain robustness,and put forward a kind of gaussian weighted Local Binary Pattern(GWLBP),according to the characteristic of gaussian function,the weighted center pixel,can better to express facial features.Experiments show that this algorithm has some advantages.(3)The algorithm in this paper,first on the Windows platform simulation experiments,using OpenCV(Open Source Compute Vision),is compiled by using Visual Studio 2015,after the completion of the algorithm,using JNI mechanism of Android,algorithm transplantation,and then use the Android Studio to the preparation of the application,has realized the one who has a human face image acquisition,face detection and user recognition,and other functions of human face recognition system.