The Design And Implementation Of Face Recognition System Base On Linux

Face recognition technology,as one of biometric technology has broad applicationprospects. Face recognition technology combines computer image processingtechnology and bio-statistical theory, widely applied to the access control system,Access Control systems, computer security, face photo search, visitor registration, ATMintelligent video alarm system, intelligent alarm prison system, and so on.This paper studies the AdaBoost algorithm for face detection based on principalcomponent analysis method for face recognition. In the Linux environment, in theembedded development board, has realized the Adaboost algorithm and face recognitionmethod, realizes the real-time face detection and recognition.This paper completes the embedded Linux platform. From the cross developmentenvironment, Bootloader Linux kernel transplantation, transplantation, root file systemproduction and transplantation, until the device driver of the transplantation process,established the embedded software development platform.The video camera image acquisition, in the embedded development boardTiny6410to achieve the image of human face detection and feature based facerecognition. This paper analyzes the Video4Linux camera based video acquisitionmethod, realizes the image acquisition module, for face detection and recognition withimage source; using Adaboost algorithm for face detection method, has realized facedetection module, laid the foundation for face recognition. Analysis of the feature basedface recognition principle, realize the face recognition module. Finally the applicationof QT design of face recognition system.The measured results show that, ARM11platform based on embedded Linux facerecognition system, in the simple light and background. Correct rate is high. Real timesex is strong, easy to carry. The work has a certain reference value in face detection andrecognition technology of embedded platform.