The Research And Application Of Virtual Model Technique Of 3D Facial Reconstruction

Facial reconstruction is very important to ethnics research, individual authentication and orthopedic operation. It is widely studied and applied in forensic medicine, physical anthropology and palaeoanthropology. In old days, 2D manual paint and 3D manual sculpture of estimating the appearance of the individual from his or her skull are extremely expensive and time-consuming. Normally a single facial estimate is produced and publicized. Nowadays, computer technology is used in reconstruction and it promotes reconstruction technology. On account of absence of mass CT images of whole head, all reconstruction methods today are based on thickness of soft tissue. These methods include acquiring data points on face from those points representing tissue thickness on skull or adjusting digital face to fit skull. All these methods are troublesome and need artificial operation.The development of computer graphics and medical image provide not only scientific visualization platform, but also a great capacity database for virtual faces and skulls. This thesis makes research on mass CT images of whole head. Firstly, these images are pretreated. Then, facial and skull's contours are exactly extracted and imitated. Lastly, relation between the model of face and skull is searched using artificial neural networks, and at the same time, many factors are taken into account, such as age, race and body fat content. Virtual model of 3D facial reconstruction techniques not only has important scientific and applied value to the research of forensic medicine, physical anthropology and palaeoanthropology, but also be useful to identity discrimination and forecast to facial operation. It has important academic and practical value to science research and life.