BRDF approximation and estimation for Augmented Reality

In Augmented Reality applications it is important to have a good description of the surfaces of real objects if a consistent shading between real and virtual object is required. If such a description of a surface is not available it has to be estimated or approximated. In this paper several methods are presented that deal with the bi-directional reflectance distribution function (BRDF) approximation in Augmented Reality. Of course an important thing to discuss is whether the applications we present work in real-time and compute real looking results. Different methods can be used to achieve these goals. All of the methods presented work via image based lighting. Some require a 3D polygonal mesh representation of the object, for which the BRDF shall be approximated. Some methods estimate the BRDF parameters via error values and provide results at each iteration.