基于隐马尔可夫模型(HMM)的人脸表情识别

Facial Expression Recognition Based on Hidden Markov Model

人脸表情识别是目前的研究热点。文中介绍了人脸表情识别的过程,给出了基于隐马尔可夫模型(HMM)的人脸表情识别方法。通过分析人脸表情的变化情况,利用二维离散余弦变换(2D-DCT)提取脸部表情特征,经过大样本训练构建HMM模型来识别图像中的人脸表情。实验结果表明该方法是一种高效的面部表情识别方法。

At present the facial expression recognition is a hot spot. The process of facial expression recognition is introduced and the method of facial expression recognition based on ttMMis discussed in this paper. Upon the analysis of the movement of facial expression, this paper describes a HMMapproach for facial expression recognition, which uses 2D-DCT coefficients as the features and constructs a HMMmodel for recognizing the facial expression. Experiment results indicate that the proposed approach is an effective method for recognizing facial expression.