摘要
针对静态人脸表情识别方法的不足,提出了一种改进的基于运动特征的动态人脸表情识别方法。以表情视频序列为研究对象,提出了基于相位形式表示脸部运动特征,处理这些运动特征并组成时序特征序列,最后将其输入到改进的高斯混合隐马尔可夫模型进行训练和测试,分析识别6种基本的面部表情。基于改进的算法,实现了一个动态面部表情识别实验系统,实验结果表明该方法简化了计算,减少了矢量量化误差。
An improved dynamic facial expression recognition method based on motion feature is proposed. Firstly, the motion features are extracted from some special facial expression regions based on video sequence, and described as phase form and then constituted to eigen-sequences. Secondly, Gaussian of Mixture Hidden Markov Model is used to learn and test these eigen-sequences, and the six universal facial expressions are recognized. An experimental system is developed based on the algorithm. The experimental results show that the computing time and the error of vector quantization is reduced, while the classification efficiency is improved.
出处
《工程图学学报》
CSCD
北大核心
2007年第5期56-61,共6页
Journal of Engineering Graphics
基金
国家自然科学基金资助项目(60573079)
关键词
计算机应用
表情识别
高斯混合隐马尔可夫模型
运动特征
computer application
facial expression recognition
Gaussian of Mixture Hidden Markov Model
motion feature
