论文部分内容阅读
开发了一种基于机器人视觉的表情识别系统。系统首先基于人脸肤色及Gabor纹理特征精确定位人脸区域,其次用Gabor小波变换提取表情特征并设计基于支持向量机的表情分类器,最终实现对7种人脸基本表情的识别。实验表明,该方法在室内环境下能够实时地识别表情并具有较高的识别成功率。
A facial expression recognition system based on robot vision is developed. Firstly, the system locates the face area accurately based on the complexion of human face and Gabor texture. Secondly, the facial features are extracted by Gabor wavelet transform and the facial expression classifier based on SVM is designed. Finally, the recognition of the basic expressions of the seven human faces is realized. Experiments show that this method can recognize expression in real time and has a high recognition success rate in indoor environment.