基于人工情感的脸部表情识别的研究  

Facial Expression Recognition Based on Artifical Emotion

作　　者：郭凤英[1] 王万森[1] 

机构地区：[1]首都师范大学信息工程学院,北京100037

出　　处：《计算机仿真》2006年第4期204-207,共4页Computer Simulation

摘　　要：脸部特征的提取和各种特征所代表的表情状态是识别是脸部表情识别过程中的重要步骤。该文研究了脸部表情识别方面的一些进展,概括论述了面部表情识别的主要研究内容,以及图像预处理,特征提取,图像识别和表情识别的主要算法和方法。研究表明,在正面人脸和无背景时识别效果较好;在复杂背景下,或人脸被部分遮挡或角度偏转角度较大时,识别效果不好,它的识别效果离实用还有较大的距离,还需要进一步研究。The extracting of facial feature and the facial expression state represented by all kinds of facial feature are important steps in facial expression recognition. This paper has summarized some progress in facial expression in recent years which discusses summarily the main research contents of facial expression recognition, as well as the main methods for image pretreatment,feature extracting,image recognition,and expression recognition. Study shows, when the image is frontal face or without environment, the effect of recognition is much better. In complicated environment,or when the face is bloted out or having big deflection angle,the effect of recognition is bad. So,the effect of facial expression recognition is far from practical use and it still needs deep research.

关 键 词：人工情感 脸部特征提取 脸部表情识别 神经网络 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]