Research And Design On A Embedded System For Human Face Detection And Recognition

Face recognition,especially the computer technology of using the analysis and comparison of human visual characteristic information for identification. Face recognition is also one of the most popular topics of the pattern recognition and artificial intelligence. It’s widely used in access control, judicial application, electronic commerce and video monitor. Most of the traditional high-performance automatic identification systems are based on the application of PC architecture. However, in practical application, automatic identification systems need more portable and more easy-to-use. Especially in recent years, with the development of smart household and the rapid development of embedded system, the importance of embedded intelligent video monitoring system is getting more and more highlighted. It combines the advantages of the embedded equipment, face automatic detection and automatic face recognition systems. It can be widely used kinds of entrance guard systems, the family security, intelligent villages and so on.The essay does a research and accomplishes a embedded face recognition system which is based on OpenCV. This essay firstly introduces some currently popular methods of face detection and recognition. Then it deeply discusses the Viola-Jones face test and PCA face recognition method in OpenCV. Through the tool functions provided by the openCV and by using the face samples which are collected to train the facade classifier. By customizing Linux kernel and file system, using the GT2400development platform to build the embedded Linux operating system, transplanting OpenCV library into the embedded operating system. In hardware, build a human face image acquisition deice with ZC301video acquisition device and GT2440development platform. Then build a face detection identification system through the way of connecting server by Ethernet. In software, adopt multithreading, modular designs. Accomplish the module designs of the video frame acquisition module, the face image frame filter module, the embedded system communication module and face recognition module through server.At the end of the essay, it summaries what the system design has done. Then it points out the defect of the system and put forward what can be improved.