Face Detection System Bacede On Static Images

Face detection means: from the source image to detect the ordinary face by use of modern digital information, including identify human face location and occupies space. Correctly detect the human face, is the precondition of face recognition, intelligent image processing. At the same time, the test technique can in human-computer interaction, security monitoring, based on image retrieval, which played a very big role application scenario.This paper expounds the development of face detection research and application at home and abroad. In recent years, with the development of information technology, face detection began as an independent subject, by the researchers’ attention. Today, face detection application scenario has been far beyond the scope of face recognition system, for the study of this technology has experienced from simple to complex, from still images to detect the development of real-time video detection, especially in recent based on Haar features face Detection years there, and truly on the PC for rapid detection of human faces. With the rapid development of the human face detection, visual applications in other fields also had a lot of similar applications, a variety of theories and methods in the development of face detection technology, For intelligent monitoring, license plate recognition and other machine vision applications, have a strong reference value.This paper reviews the current development of domestic and foreign recognition technology research and application, as well as difficulties in analysis, introduction to the mainstream face detection method, discuss the advantages and disadvantages of each method.This paper studies the human face detection method based on the HAAR features., Consists of two parts: haar feature description, and Adaboost algorithm. AdaBoost algorithm is an important mechanism for integrating machine learning : Analysis of the multiple weak classifier integration method. In the text, but also details the principles and methods of training, training classifiers Adaboost classifiers detailed design process, and gives training program source code.In this paper, Face detection system is designed, based on openvpn source image library. In OpenCV open source image library provides a lot of basic data types and system construction method, we use this set of sophisticated image processing program framework, with VS2010 on windows, build a face detection system for static images. Finally, Test to verify the success rate and error rate and detection rate, Choose typical image was tested. Test results show that in this paper, to achieve 100% detection through no fault detection and missed cases, testing to good effect.