Research On Synthesizing Realistic 3D Face

The human face is the most familiar organ to us, but it is a difficult problem in the area of computer graphics to synthesize realistic human face. Since the pioneering work of Parke F.I. in 1972, the realistic facial synthesis is still a subject with great challenge and worth deeply researching to researchers in the computer graphics field. The hunman face modeling and its expression animation can be widely used in movie, computer game, visual telephone and human-computer interface.This paper focuses on the field of facial synthesis. The research work mainly includes the following aspects:(1) According to the FAP in the standard of MPEG-4, we define some feature points in the human face based on the characteristics of human face.(2) The paper studies the method of deriving the 3D coordinate of the feature point from the 2D images. With this feature points, we get a given face model by modifying the generic face model. The modification of generic face model includes two steps, one is whole modification, and the other is part modification.(3) We use image pyramid method to merge the boundary of the original texture image. It must build a Gaussian pyramid and a Laplacian pyramid. The process of building a pyramid is to produce image sequence with different scales.(4) Finally, we use the technique of cylinder texture mapping to generate a realistic 3D face.(5) Besides theory research, we also built a 3D face synthesis system, using Visual.Net and OpenGL on the PC platform. System use the method mentioned in paper, the result of the test has verified its feasibility.