计算机应用与软件Issue(3)：171-174,210,5.DOI:10.3969/j.issn.1000-386x.2014.03.045
基于增量学习和 ASM 的人脸表情分析与识别
FACIAL EXPRESSION ANALYSIS AND RECOGNITION BASED ON INCREMENTAL LEARNING AND ACTIVE SHAPE MODEL
梁雪梅1
作者信息
- 1. 重庆电子工程职业学院计算机学院 重庆 401331
- 折叠
摘要
Abstract
Active shape model (ASM)is a parameterisation-based statistical model,which is mainly used in image feature points extrac-tion and image segmentation.An improved new method is proposed which uses ASMto locate the facial features based on analysing the insuf-ficiency of traditional method.By using incremental learning PCA,this method is able to effectively resolve the factors of model matching fail-ure and the effect of the image to be tested,etc.,and to update the texture model on training set at the same time.Moreover,the improved method is used for face expression analysis and recognition,and at last the SVMis used to set up the expression classifier.Experimental re-sults show that the improved method can effectively improve the locating accuracy of facial feature points and also improve the expression rec-ognition rate meanwhile.关键词
主动形状模型/特征提取/PCA/增量学习/纹理模型/表情识别/SVMKey words
Active shape model/Feature extraction/PCA/Incremental learning/Texture model/Expression recognition/SVM分类
信息技术与安全科学引用本文复制引用
梁雪梅..基于增量学习和 ASM 的人脸表情分析与识别[J].计算机应用与软件,2014,(3):171-174,210,5.
