【摘要】 针对多姿态人脸识别问题,提出了一种三维形变模型和部件技术相结合的人脸识别方法．该方法提取人脸三维部件作为识别特征,在一定程度上克服了人脸姿态变化对识别的影响．在识别过程中结合人脸的局部特征和全局特征,根据单个部件的识别率确定其在整体分类中的权值,基于整脸信息进行识别,进一步改善了识别效果．实验结果证明,该方法在训练库中只有1张正面照片时,对多姿态人脸识别具有很好的识别效果．更多还原

【Abstract】 In this paper,a novel method to pose variant face recognition that combines two recent advances of component-based face recognition and 3D morphable model is presented.In this method,3D components are extracted as the feature of face recognition.Because of the Shape information,it reduces the effect of pose to face recognition.For classification,we combine the local feature and global feature of face and the whole face are used as input to the final classifier,where each component is verified its weight based on its recognition rate in final classifier.Experiment results show that the methods is robust to pose invariant face recognition with only one image of each person in the gallery.更多还原