Face Recognition In Mobile Environment

Face recognition is the use of facial feature information biometric identity authentication technology,as opposed to the traditional identity authentication technology,face recognition is universal,stability,not forgetting other advantages.Under controlled conditions,the existing face recognition systems generally have good recognition performance.B ut in the mobile environment,environmental illumination,facial expressions,posture and partial occlusion and other factors become more pronounced.Thus,face recognition technology will be applied to the mobile environment is still faced with many probl ems and challenges.In this paper,Aiming at mobile environment problems such as illumination,facial expression,posture,and local shelter,from the human face feature extraction and recognition scheme in mobile environment two aspects were studied,and to achieve a face recognition system based on Android platform.The main work is as follows:1 Study face recognition feature extraction methods,we propose a partial description of Weber(Weber Local Descriptor,WLD)and gradient-based face(Gradientface)feature extraction method(GWLD).GWLD face image feature extraction differential incentive information and gradient face information,and feature two parts information fusion,get a better judgment of facial features.Experiments show that,GWLD characteristics compared to other features have certain advantages in the recognition rate and robustness.2 Research the illumination problem of mobile environments,combined with smart phone features,face recognition scheme is proposed based on the mobile environment.The program mainly consists of a screen-based technology and a fill light illumination sensor-based pretreatment process,the screen fill light ambient lighting technology used in quite a bad situation to improve the clarity of the face image,based on the pre-light sensor process for improving the recognition process in the preferred case the speed of light.The establishment of a human face database mobile environment,including indoor lighting,outdoor light,night lights,sunglasses block and screen fill light five subsets.Experiment s show that light up the screen-based technology can significantly improve the recognition rate in the absence of external light situations;based pretreatment technology light sensor under normal lighting ca n improve recognition speed.3 Based on the results of the mobile environment recog nition,to achieve a lightweight Android platfor m based face recognition system.In this paper,this system architecture,design,function and features a detailed description and analysis.