The Facial Expression Recognition Technology Based on D-S Theory with Weighting Coefficients

作     者：MA Xirong, WANG Rong , LIANG Jinglian College of Computer and Information Engineering, Tianjin Normal University, Tianjin 300387, China 

出 版 物：《International Journal of Systems and Control》 (国际系统与控制(英文版))

年 卷 期：2008年第1期

页      面：35-41页

学科分类：1305[艺术学-设计学（可授艺术学、工学学位）] 13[艺术学] 081104[工学-模式识别与智能系统] 08[工学] 0804[工学-仪器科学与技术] 081101[工学-控制理论与控制工程] 0811[工学-控制科学与工程] 

基　　金：the Natural Science Foundation of Tianjin under No.08JCYBJC13700 the Science and Youth Foundation of Tianjin Normal University (No.52LJ88) 

主　　题：D-S theory with weighted facial expression recognition Affective Computing Information Fusion 

摘      要：This paper proposes a kind of facial expression recognition technology based on the D-S theory with weighting coefficient of information fusion, in the background of Affective Computing theory, moreover design and implement the system FERFIF. In the part of facial expression recognition, the training for the classification uses JAFFE expression library. First, using color matching and brightness match, this method divides face image into the eyes and mouth images, then identifies facial expression respectively by trained eyes SVM classifier and mouth SVM classifier, in the end, fuses the identification using D-S theory. Because the eyes and mouth expressed different amount of information, in the integration, the two groups’ data were weighted. The experimental results show that recognizing the two parts of image, in terms of training or recognition, the data dimension greatly reduced and increased efficiency. Moreover the improvement of weighting data makes better result than before in recognition rate.