Face Detection Based On Adaboost Algorithm

In recently years,face detection has became a more popular research topics in computer pattern recognition.It can be widely applied to such fields as criminal archive administration,personal indetification,visual communication,etc.The task of face detection is to detect human face from the complex background image.It's the most important part in face recognition.However,the result of face detection is affectde by image background, lighting,facial expression or head posture of image and so on. These problem makes face detection is a very hard task. In this thesis,the research work of this paper mainly Includes the following several aspects:1. Using face detection method based on Adaboost learning algorithm,which select few key haar-like features from a large set of features, to build a robust cascade classifier to detect face. Because original method can't detect profile face. Inspired by the Adaboost algorithm, we train profile face samples by Adaboost algorithm to solve the proble of detect profile face.2. We select YCbCr space, and use gaussian model to extraction skin color. We proposed a method that calibration face in skin color image based on contour. The experimental results show that the new method effect if remarkable.3. Finaly, i summarize shortage of two method when they were used in detecting face, then, a novel face detection method combined skin color detection and Adaboost algorithm is proposed in this paper. The experimental results show that the new scheme is able to detect faces with high detection rate and low false acceptance rate and can save time.