【摘要】 针对三维人脸数据中可能存在的尖点、孔洞以及姿态变化等因素会造成人脸样本类内差异增加从而影响人脸识别系统精度的问题,提出一种三维人脸数据预处理方法.实验表明,通过切割、数据规整、平滑和姿态矫正使原来散乱的三维人脸点云数据变为规整的人脸点云数据,并可弱化噪声和冗余信息的干扰降低人脸样本的类内差异.更多还原

【Abstract】 In 3D face data,these factors existence such as the of the cusps,holes,pose vary and so on,will cause the class differences increase of face and effect the the precision of face recognition system. According to this problem,a novel 3D face data pre-processing scheme is proposed. The experiment result showed that the scheme could transform the scattered point cloud data into norm data,weaken noise and decrease the within-class difference by segment,norming,smoothing,and pose correction.更多还原