Model-based Face Capture from Orthogonal Images

From a survey of techniques, a novel method was implemented for creating photo-realistic 3D facial models from 2D images of a subject. Starting from two orthogonal views, a userassisted procedure was employed to recover 3D coordinates of a sparse set of chosen locations on the subject's face. A scattered data interpolation algorithm was then used to deform a generic face mesh to t the particular geometry of the subject's face. A model texture map was created by combining both views. Using this technique, it is possible to generate highly realistic facial models. These models can be used for diverse applications including low-bandwidth video conferencing, computer animation, and multiple-view face recognition.