Facial expression recognition is widely used in the fields of Man-Machine Interaction, intelligent robot, safe driving and clinical medicine. The main research contents in this paper are as follows:1. The research background and the current research situation of the facial expression recognition are illustrated. The main technologies of facial expression feature extraction and classification in the facial expression recognition are generally described.2. A new method of eye location is proposed. It detects a roughly eye area with Adaboost in the first. Then it reduces the area with rectangular mask and elliptical mask. At last eye locations are obtained by computing the distribution centre of pixels.3. Considering the three deficiencies of local binary pattern (LBP), Center-based binary pattern (CBP) is adopted at the stage of feature extraction. Compared with LBP, CBP, which considers the center pixel, reduces the feature dimensions and changes the sign function. Experiment results show that the method has a fast speed and a good ability to overcome the noise.4. Center nearest neighbor (CNN) classifier and two-versus-two multi-class SVM classifier are adopted to classify the expressions. Compared with the CNN and the other methods, the result of experimentation shows that two-versus-two multi-class SVM classifier has a fast classification speed and a high performance. |