A 3D deformable model constrained by anthropometric knowledge for computerized facial reconstructions

In this paper, a three-dimensional computerized facial reconstruction method from three-dimensional Magnetic Resonance Image (3D-MRI) of human heads is presented. It based on a model deformable approach constrained by the knowledge of soft tissues thicknesses, at a certain number of characteristic landmarks. The facial reconstruction process is divided into two main stages. An automatic initialization step of the deformable model is based on a non-linear registration technique guided by B-Splines Free Form Deformations model. In the second step, the initial skin, resulting from the transformation computed previously, is refined by means of a three-dimensional deformable model based on simplexes meshes that is attracted by some characteristic landmarks previously computed by a calculation of Mean and Gaussian curvatures. The evolution of the deformable model is done according to different internal and external forces among which the Gradient Vector Flow force and a balloon force. This method has been tested on thirteen sets of skull/skin data, extracted from 3D-MRI of Image heads of children and adults.