摘要
针对包含表情信息的静态图像,提出基于皮肤检测和SVM的人脸表情识别算法。首先根据先验知识,并使用皮肤检测和积分投影相结合定位眉毛眼睛区域和嘴巴区域,自动分割出表情子区域。接着,对分割出的表情子区域进行Gabor小波特征提取,在利用Fisher线性判别对特征进行降维,去除冗余和相关。最后利用支持向量机对人脸表情进行分类。用该算法在日本表情数据库上进行测试,获得了较高的识别准确率。证明了该算法的有效性。
A facial recognition algorithm based on Skin detecting and SVM to still image containing expression Information was introduced. Firstly, skin detecting algorithm combined with projection to locate the eye region and the mouth region, which can segment the expression sub-regions automatically. Secondly, features of the expression sub-regions were extracted by Gabor wavelet transformation and then effective Gabor expression features were selected by Fisher Linear Discriminat （FLD） to deduce the dimension and redundancy of the features. Finally, the features were sent to Support Vector Machine （SVM） to classify the different expressions. The algorithm was tested on Japanese female expression database. It can get a high precision of recognition. The feasibility of this method has been verified by experiments.
出处
《电子设计工程》
2011年第24期150-153,共4页
Electronic Design Engineering
基金
西北工业大学2011年度研究生创业种子基金(Z2011090)
关键词
亮度检测
表情特征提取
FISHER线性判别
支持向量机
brightness detecting
expression feature extraction
Fisher Linear Discriminant （FLD） analysis
Support Vector Machine （SVM）
