The Study Of Multi-expression Classification Algorithm Based On Adaboost And Mutual Independent Feature

Facial expression can show state of mind from the side and embody human behavior information, which provide reliable information for us to study the state of mind. Facial expression recognition which is developing quickly based on facial emotional information.Currently there are many main methods used for facial expression classification, such as Euclidean distance, Support vector machine (SVM), Neural network(NN), Hidden Markov(HMM). Adaboost and Linear discriminate analysis (LDA), etc.Adaboost algorithm which owns higher speed and higher detection rate has been successfully applied in the field of face detection. Most of facial expression changes exist in eyes and mouth, so features of eyes and mouth are treated as mutual independent elements. The method can greatly reduce redundancy and improve the speed of training threshold values. However, if we want to use the algorithm to facial multi-expression classification, key problem is to train a weak classifier threshold value accurately and fast. So we improve training samples, and negative samples are proposed in this paper. Threshold value of weak classifier is the most crucial part. Positive samples contain only images of eyes or mouth, while negative samples are removed eyes and mouth. So the multi-expression classification algorithm which is based on Adaboost and mutual independent feature is proposed. Experimental results prove that false recognition rate is almost close to zero.