A New Shape from Shading Approach for Specular Surfaces

Shape recovery is a basic problem in computer vision. Shape from shading (SFS) is an approach to get the 3D shape from a single shading image. Diffuse model is usually used to approximate the surface reflectance property. For specular surfaces, however, it is not suitable. In this paper, we propose a new SFS approach for specular surfaces. The Blinn-Phong reflectance model is applied to characterize the specular reflection property. The image irradiance equation for specular surfaces is obtained under the assumptions that the camera performs an orthographic projection and its direction is the same as the light source. Then, it is formulated as an Eikonal PDE which includes the shape of the surfaces. The viscosity solution of the resulting PDE is approximated by using the high-order Godunov fast sweeping method. Experiments are performed on both sphere and vase images and the results show the efficiency of the proposed approach.