Fast Face Detection System Design And Implementation

This thesis focuses on how to detect faces in gray or color images at high speed and accuracy. Face detection is a basic and important research theme in the filed of Pattern Recognition and Computer Vision. Face detection has the application in a wide range of fields such as automatic face recognition, visual surveillance and content-based retrieval. However, face detection is a most challenging task because of the complicated pattern and frangibility of human face. So, most face detection methods have the weakness of large computation, low efficiency and unstable, etc.Based on the technology of real-time face detection presented by Dr. Paul Viola[l,2], some extension to the technology are made and a rapid face detection system is designed and realized. The experiment results show that the extension to Viola's technology is quite successful. The face detection system achieves a high hit-rate and low false-alarm-rate, especially, the detecting speed of the system is very high and almost attain the criterion of real-time.The rapid face detection system is divided into two parts, training(or learning) subsystem and detecting subsystem. The focus is on the training subsystem. During describing the design process and algorithms, many important points are singled out and analyzed in particular. Groups of experiment results are compared with previous system at the end of the thesis.