一种自动鲁棒的三维人脸重建方法

An Automatic and Robust Approach to 3D Face Reconstruction

三维人脸模型已经广泛应用到视频电话、视频会议、影视制作、电脑游戏、人脸识别等多个领域。目前三维人脸建模一般使用多幅图像,且要求表情中性。本文提出了基于正、侧面任意表情三维人脸重建方法。首先对二维图像中的人脸进行特征提取,然后基于三维人脸统计模型,通过缩放、平移、旋转等方法,及全局和局部匹配,获得特定的三维人脸。基于二维图像中的人脸纹理信息,通过纹理映射,获得完整的三维人脸。通过对大量实际二维人脸图像的三维人脸重建,证实了该方法的有效性和鲁棒性。

3D facial model have been widely used in many fields, such as video telephone, video conference, movie making, PC game, face recognition, etc. Multiple images are usually used in 3D facial modeling and faces in images must be natural expression. This paper proposes an approach to using frontal and lateral 2D image, to reconstruct a 3D face. First of all, facial features are extracted from the face in the 2D image, which contain many useful information for 3D reconstruction. Then a generic 3D face model is adopted to fit a special face in the 2D image by means of scaling, shifting and rotating. Then, facial texture is obtained from the image. After mapping the 3D face model co-ordinates into the texture space, texture co-ordinates of the 3D face model can be generated. Experimental results show that the developed method to reconstruct 3D face is effective and robust.