Facial Expression Recognition Based on Active Shape Model and Gabor Wavelet

摘要

Based on the combining ASM (active shape model) and Gabor wavelet, this paper presents a facial expression recognition algorithm named ASM_GW, used to achieve significant reduction of facial expression feature dimension under the condition of ensuring some facial expression recognition ratio. First, the principle of locating feature points by use of ASM is given. Second, how to create the feature vector corresponding to the feature points is explained based on Gabor wavelet. Third, PCA (principal component analysis) plus LDA (linear discriminant analysis) is used to finish the feature dimension reduction and the facial expression classification. At last simulation experiment of the algorithm is performed by use of JAFFE facial expression database to demonstrate the effectiveness of the algorithm.