Haar Feature Based And Multi-Feature Utilized Rapid Face Detection System

Face recognition can be widely used in such fields as personal identification, information security, e-bussiness, video surveillance.multi-media, etc. With the demanding of society and bussiness, developing of computer hardware and software and progress of face recognition research, face detection become more and more important.It's a key of face recognition application, especially in real world.The research of face detection advanced from simple image process to complex real-time video process.Haar feature based face detection made PC can see human face via camera.Our experiments show that the Haar feature based face detection algorithm can truly detect face rapidly, but face detection rate and false negatives rate is not satisfying, especially videos in normal environment.Other face features, such as skin color, contour and edge energe are used in our face detection system, skin model, contour selection, canny edge pruning are utilized in order to improve the system performance, raise face detection rate and cut down false negatives rate.In this paper, a face detection system using DirectX technology may be useful for other computer vision system, such as face recognition and vehicle plate recognition.