Abstract: Based on analysis of the pro-existing face pose estimation methods, a new 3D face pose estimation method based on Active Appearance Model（AAM） and T-structure is proposed. A set of AAM models can be obtained after training different poses faces. Then, the objective face is matched with the set of AAM models, to choose the optimum model to accurate position of the face characteristic points. The T-structure is built with the two eyes and the mouth, which is used to estimate the face pose. The experiments show that the method can be adopted to large rotation angles, and can reach a good accuracy of 3D face pose estimation.

Key words: Active Appearance Model（AAM）, face pose estimation, characteristic points location, T-structure, face detection

摘要： 在分析已有的人脸姿态估计方法基础上，提出了一种基于主动表观模型（AAM）和T型结构的人脸3D姿态估计方法。对多姿态的人脸样本进行训练，得到多姿态的AAM模板集；利用训练得到的多姿态的AAM模板集进行最佳模板匹配，并对人脸的特征点进行精确定位；用人脸的双眼和嘴部构建T型模型，进行人脸3D姿态的参数估计。实验结果表明，该方法能适应较大的姿态旋转角度，并具有良好的姿态估计精度。

关键词: 主动表观模型, 人脸姿态估计, 特征点定位, T形结构, 人脸检测