Realistic 3D reconstruction of the human teeth using shape from shading with shape priors

Dentistry usually requires accurate 3-D representation of the teeth and jaw for diagnostic and treatment purposes. Photogrammetry can offer a flexible, cost-effective solution in that regard. In this paper, we develop an approach for realistic 3D reconstruction of the human teeth using shape from shading with statistical shape priors. Our work has addressed several challenges including near illumination, camera perspective projection, while taking into account the deviation from the simplifying Lambertian assumption. We use the Oren-Nayar reflectance model for diffuse rough surfaces with the roughness parameter being physically measured by an optical surface profiler. Our formulation exploits the shape priors as extracted from a set of training CT scans of real human teeth. Our experiments provide quantitative metric results for the proposed approach. As compared to state-of-art SFS approaches, our approach is able to recover geometric details of a tooth occlusal surface.