Face Expression Recognition Base On DSP

Expression information important to people exchange, with the continuous developmentof artificial intelligence and pattern recognition, an important technology-Facial expressionrecognition in human-computer intelligent interaction more and more attention has been paid.Facial expression recognition is the computer on the expression of the face feature extractionand classification process it so that the computer can be informed of the person’s face, andthen infer the mental state of the person, in order to achieve the advanced intelligentinteraction between man and machine. In this thesis, facial expression classification for facialexpression classification and identification of the application technology.This expression recognition system including face detection, facial feature extractio n,feature selection and expression classification. The design is mainly based on the BF533DSPchip, running clinux system on BF533chip, achieve facial expression recognitiontechnology. First used AdaBoost algorithm as face detection and eye localization algorithm,Normalization processing and the image size, the angle, position and gray value; then usedPCA add LDA extracted as facial expression features; Finally, used National TaiwanUniversity, Dr. Lin Zhiren development and design LIBSVM, by support vector productalgorithm expression classification and recognition...