Surface reflectance characterization by statistical tools

The classification of surface reflectance functions as diffuse, specular, and glossy has been introduced by Heckbert more than two decades ago. Many rendering algorithms are dependent on such a classification, as different kinds of light transport will be handled by specialized methods, for example, caustics require specular bounce or refraction. As the surface reflectance models are more and more rich and descriptive including those based on measured data, it has not been possible to keep such a characterization simple. Each surface reflectance model is mostly handled separately, or alternatively, the rendering algorithm restricts itself to the use of some subset of reflectance models. We provide a general characterization for arbitrary surface reflectance representation by means of statistical tools. We demonstrate by rendered images using Matusik's BRDF data sets for two environment maps and two 3D objects (sphere and Utah teapot) that there is even a visible perceptual correspondence to the proposed surface reflectance characterization, when we use monochromatic surface reflectance and the albedo is normalized for rendering images to equalize perceived brightness. The proposed characterization is intended to be used to optimize rendering algorithms.