Lip Features Basic Expressions Of Classification Algorithm

Facial expression recognition has been one of the most popular and difficultproblems in the field of pattern recognition, computer vision, and artificial intelligencein recent years. Expression recognition technology is of great significance for intelligenthuman-computer interaction technology and has great potential value in the field ofmedical applications, entertainment, distance-education and so on. Studies have shownthat lip region has the greatest impact on the expression when people make differentexpressions. Besides, study on lip region is valuable to lip-reading technology.Therefore, this paper focused on the lip region of face, studying and achievingexpression feature extraction and classification of the lip region.Firstly, the preprocessing of face images is necessary in order to obtain the lipregion images with the same size. The next step is to extract geometric features andGabor features, including direct geometric features and indirect geometric features.Uniform sampling is useful before Gabor feature extraction. Besides, Adaboostalgorithm is used for feature dimension reduction. Expressions classification in the lipregion is the last step by using support vector machine. The whole process isimplemented in the JAFFE library.Comparison the method proposed in this paper and traditional expressionrecognition algorithm in the perspective of recognition rate is made in the last part ofthis paper. Advantages and effectiveness of the methods in this paper are proved by theexperiment results.