3D facial reconstruction from a single 2D rasterstereography image

This paper presents a new 3D reconstruction approach for the human face, inspired from raster-stereography, a non-intrusive technique usually employed in vertebral imaging. The method proposes depth estimation of the facial topography based on raster grid deformations, as well as a parameter-free way to obtain the reconstructed topographic surface. An example application of the depth map reinforces the correctness of the approach.