Research Of Human Facial Expression Recognition

Facial Expression Recognition (FER) has great potential value, which is a set of physiology, psychology, computer science and other disciplines of cross-subject. A typical Facial Expression Recognition system usually contains the following three Parts:facial expression database detection, facial expression feature extraction and classification of facial expression.First of all, in facial expression database detection, the view of completely open and standard expression database, JAFFE face database is selected as the training and testing of expression database. By bilinear interpolation expression image preprocessing is done to reduce the dimension for feature extraction and classification. Secondly, to solve the feature extraction problem, algorithm is proposed in this paper, which adopts1CA (Independent Component Analysis) to extract feature. ICA is used to extract statistical independent feature which is a higher-order statistical methods, emphasizing on the independence between the various components, and it can describe the essential characteristics of the image. In this paper, the new method is quadratic ICA, which can improve the independence of neural network variables to enhance the accuracy of pattern recognition. SVM (Support Vector Machine) is a widely used method in pattern recognition and it is used as the classifier. By POS (Particle Swarm Optimization), SVM parameters the best parameters are obtained. Finally, PNN (Probabilistic Neural Networks) is used to classify the image feature. In classification of SVM and PNN, experimental results show the quadratic ICA is an effective conduct of facial expression recognition.