查看文章

In this paper, we present a new semiautomatic method to reconstruct 3D facial model for animation from two orthogonal pictures taken from front and side views. The method is based on extracting the hair and face outlines and detecting interior features in the region of mouth, eyes, etc. We show how to use structured snakes for extracting the profile boundaries and facial features. Then DFFD, a deformation process, is used for modifying a predefined or generic 3D head model to produce the individualized head. Texture mapping based on cylindrical projection is employed using a composed image from the two images. The reconstructed 3D face can be animated in our facial animation system.