Face Reconstruction Based on Camera-Projector System

With the coming of information age and the development of computer science, digitalization of human information is becoming more and more important. Face as the most important characteristic of human has become the most challenging subject of computer 3D reconstruction. The paper proposes a method by which face 3D model is created from 2D image sequences directly but not from the common techniques using face general model. In the paper the structure illumination is generated by the projector and digital camera system and the reconstruction of different faces is based on the strict theory of the digital photogrammetry, computer vision and image processing pattern recognition. Because face is lack of the applicable texture for matching, the characteristic texture generated by the structure illumination is added onto the surface of human face. After the characteristic texture is extracted from images and is matched well, 3D coordinates of the characteristic texture can be calculated out by the space forward intersection. Finally, the face model is acquired by connecting all neighbour space points. The method of face reconstruction proposed in the paper is flexible, effective and practical, which is confirmed by the results of the reconstructing experiments.