Realistic 3D Face Texture Mapping

Orthognathic Surgery is a medical treatment which has the high risk. With the development of computer virtual reality, how to carry out surgical planning and simulation effectively, and how to improve the rate of success have become hot research fields. Under the National Natural Science Fund Project "An orthognathic surgical planning and simulation based on 3D statistical deformation and finite element hybrid model", in this thesis, we focus on the research of CT image pre-processing, three-dimensional reconstruction, texture images generating and texture mapping. Furthermore, a prototype system for photo-realistic 3D human face mapping is developed based on above research results. The main contents of the research of this thesis are listed as follows.1. Face geometric model reconstruction. For the characteristic of two-dimensional CT image, we pre-process the image sequences by denoising, correcting and redundant data removing. The surface rendering algorithm of Marching Cubes is adopted for three-dimensional reconstruction.2. Human face Texture Images reconstruction. To overcome the shortcomings that it is imprecise to use feature line to locate the intersection of the adjacent photographic images in traditional texture image generation method, a new algorithm called "Geometric model segmentation" is introduced. First, through segmenting the geometry model of human's head, the contours of the divided model photos were obtained to cut the corresponding face photos as the boundary line. Then, the cut face photos are used to generate texture pictures without overlapping and gap between the adjacent pictures. Finally, facial texture image of full view is obtained. 3. Mapping from texture image to 3D geometric model. The cylindrical projection is adopted to construct photo-realistic 3D specific face model through the texture mapping by using of the generated texture image.4. System design and implementation. Based on the above work, the prototype system of computer photo-realistic 3D human face mapping called "VRFace3D1.0" is developed with the functionalities of CT image pre-processing, three-dimensional reconstruction, texture images generating and texture mapping.The research of the thesis laid a good foundation for the application of surgical orthopedic surgery in virtual orthopedic surgery, and it has the important significance in selecting optimal path before surgery, reducing the surgical injury and improving the rate of success.The thesis is supported by "National Natural Science Fund Project (Project Number: 6087309)" and "Northwest University Graduate Innovative Education Project (Project Number:09YZZ66)".