Research On The Application Of Face Recognition Based On ARM In Smart Access Control

Face Recognition as a “judge a book by its cover” bio intelligent recognition technology, compared with other biometric technologies, has the advantages of easy to get, easy to identify, high accuracy, intuitive good etc. Nowadays, face recognition technology has been widely used in various fields at home and abroad, especially in terms of security. In life, access control is more common security system, but the traditional access control because of its non intelligent can not effectively protect the personal safety and property safety. Therefore, in view of the face recognition advantages and the traditional access control shortcomings, this paper designs and develops face recognition intelligent access control system based on embedded technology.The system is based on embedded technology, selects lower cost ARM platform for system construction. The hardware platform chooses TQ2440 development board and V26 camera, the software platform chooses Ubuntu12.04 as the operating system. The system includes five modules: image acquisition module, user interface module, image preprocessing module, face detection module and face recognition module. Different module uses corresponding method, image acquisition uses V4L2 capture images, user interface module uses QT and Qt Creator installed build user interface, image preprocessing module uses improved histogram equalization process images, face detection module uses the Ada Boost algorithm based on Haar features detect face, face recognition module based on an improved PCA algorithm recognizes faces. Combined with five modules, the program is compiled and transplanted to ARM board.After running ARM board, the first thing establishes basic image library for recognition. The system captures different expressions and postures of each person by a USB camera connected ARM board, captures facial features to be stored in the form of images, to be generated an image library and associated data files. The process of face recognition is, capturing face image to be identified, compared with the image library. When data is matched, the screen displays related personnel information and allows to be entered, otherwise refuses. In particular, the system takes into account confounding factors such as light changes and appearance changes in special processing and analysis, to ensure experimental results well.