An efficient BRDF acquisition for glossy surface

An efficient BRDF(Bidirectional Reflectance Distribution Function) measurement of glossy surface by photometric stereo was proposed. We assume that the surface reflectance can be approximated by the sum of a Lambertian and a specular component. We define the best window in one image and classify the patches into the diffuse points and the mix ones according to their radiance values. The diffuse parameter and surface normal are recovered by the diffuse points and the darkest three pixels of four pixels. The specular reflectance parameters are recovered by the specular components of mix points which normal and diffuse parameter are known. In the last, all reflection parameters are refined by minimizing the fitting error of the reflectance model. Experiments show that the reflectance of glossy object can be recovered in a good precision by the proposed method.