Research And Implementation Of Embedded Face Recognition System Based On ARM

Face recognition is a biometric identification technology that has been used to authenticate based on human face feature information.At present,it is a hot research topic in the field of computer vision and pattern recognition.It has a wide range of applications in the field of image processing,intelligent monitoring,intelligent vehicle system and so on.Most of the current face recognition products are based on the x86 architecture,in some cases it is very inconvenient.Therefore,the development of portable embedded face recognition products has theoretical significance and practical value.A complete face recognition system includes four parts: face detection,image preprocessing,feature extraction and face recognition.In this paper,AdaBoost face detection algorithm based on Haar feature is implemented in the face detection phase.In order to determine whether the detected face is a positive face,the human eye detection algorithm based on integral projection is proposed.In the image preprocessing phase,the geometric normalization and histogram equalization are carried out to eliminate the influence of different shooting conditions on the size,location and illumination of the image.In the phase of feature extraction,PCA feature extraction method and 2DPCA feature extraction method are studied,and the two methods are compared by experiments.Because the 2DPCA method takes up a large amount of memory space,and the embedded system resources are limited,this paper adopts the method of PCA feature extraction.In the phase of face recognition,the Euclidean Distance is used to calculate the projection points of each face image in the face space to judge which face to be recognized.Finally,the Tiny210 development board is used as the hardware experimental platform,and the related program code is transplanted on it to build the embedded face recognition system.The processor is S5PV210 based on the ARM Cortex-A8 core,equipped with the Linux operating system.In order to facilitate the interaction between human and computer,a graphical interface based on Qt is designed.Then I built a small face database used to test the system.The experimental results show that the embedded face recognition system can recognize the face correctly when the illumination is more uniform.