Embedded Intelligent Face Recognition Access Control System

In recent years, with the popularity of anti-theft security devices and intelligent household products, biometric identification products into our daily life, in which the most representative biometric technology is human face. Access control system is an important application of face recognition technology can be said to have a very broad prospect, the previous platform of face recognition access control system, most are bulky PC not only has a high cost,but inconvenient to carry. However, with the emergence and rapid development of embedded technology, the problem of limited to the PC platform is being improving, the possibility of more miniature and more portable recognition products are being come ture.This paper studies for face recognition system by using embedded technology and the combined method of facial recognition technology, in view of the residential area and small population applications. It is based on embedded ARM platform and the Linux operating system, combining with the Qt framework and open source OpenCV image processing library, selecting S5pv210 as the system CPU, using Adaboost algorithm to realize face detection and blink detection, using PCA algorithm to realize face recognition.Firstly, the host Linux system to build a cross-compiler environment, compile the Linux kernel, make the file system, install the Qt and OpenCV image library, establish embedded software development required platform. Secondly, the paper analyzes the camera-based video capture processes based on Video4Linux2 framework. achieves a video image capture by using the USB camera which provides a prerequisite for face detection and recognition; realizing visitors’face detection and Activity judgment lay foundation for face recognition by using Adaboost algorithm of face detection and blink detection methods; realizing visitors’ identity judgment by using PCA feature face recognition algorithm. Finally, the performance of the system was tested, the results were analyzed.System test results show that the face recognition access control system has high accuracy and strong real-time in a simple background and lighting conditions, and it can be extended to portable face recognition device because of S5pv210 processor embedded devices in the future.