作　　者： ; ; ;

机构地区： 中国科学院研究生院

出　　处： 《计算机仿真》 2011年第11期235-237,245,共4页

摘　　要： 研究疲劳状态的表情识别问题。针对当人体处在疲惫状态与常规表情中的人脸面部特征变化不是很明显,在普通状态下的面部表情差别不大。传统的表情识别算法很难准确高效的识别。为提高识别率,提出了关联面部特征的表情识别方法。首先对人脸主要图像进行特征检测,然后建立特征集,根据特征向量之间的变化将细节特征向量进行有效联系,克服传统方法的弊端,根据微小的关联变化完成表情识别。实验结果表明,方法的能够对面部特征细微变化下的人脸图像进行有效的疲惫表情识别,提高了识别的准确度。 Study the identification of fatigue expressions.When bodies are exhausted,the facial features do not change much,and are difficult for general expressions identification algorithms to identify accurately and efficiently.This paper put forward a relational expression facial characteristics identification method.The method firstly made feature detection for main image,and then established feature set.According to the changes in feature vector,effective contacts were made between feature vectors for detail,which overcome the disadvantages of traditional methods.The experimental results show that the method can effectively identify the facial expression of exhaustion with subtle change of face images and improve the identification accuracy.

关 键 词： 疲劳表情 面部特征 关联变化

领　　域： [自动化与计算机技术] [自动化与计算机技术]