Craniofacial Surgical Planning Based on Projected Image

INVESTIGATORS

Dr. Horace H S Ip, Mr. Lijun Yin

 

BRIEF DESCRIPTION

We proposed a new scheme for generating 3D individual face model from two views of a person's image. Unlike the traditional methods such as 3D digitizer face modeling, modeling by sequence of images and some interactive techniques, we developed an approach which is based on an instantiation of a generic facial model using facial features extracted from side and front views of a person. In general, there are four stages in this scheme,

  1. Feature vertices definition of generic facial model in both view.
  2. Feature extraction for both views of the individual face.
  3. Generic facial model adjustment in both views separately, and fitting the model to the individual face in both views.
  4. Synthesis of individual 3D face model and generation of the real facial images in arbitrary views using blending texture-mapping.

 

FUNDING AGENCY

N/A

 

PUBLICATIONS

  1. Horace H S Ip and Lijun Yin, "Arbitrary Facial Views Generation From Two Orthogonal Facial Images", 1995 IEEE International Conf. on System, Man and Cybernetic, Vancouver, Canada, Oct. 1995.
  2. Horace H S Ip and Lijun Yin, "Constructing 3D Individualized Head Model From Two Orthogonal Views",the International Journal in Computer Graphics,"The Visual Computer", Vol.12, No.5, 1996. Springer-Verlag.

 

FIGURES OR IMAGES

(48kb)

Fig.1 : two views of a generic 3D face model.

(11kb)

Fig.2 : two views of a person's face.

(29kb)

Fig.3 : generated individual face model in different views.

(12kb)

Fig.4 : arbitrary facial views generated after texture-mapping.