Photometric modeling of specular and diffuse appearance

The appearance of an object can vary considerably with changes in illumination conditions. Methods have been developed to describe these differences for diffuse reflections using the Lambertian model, but little research has been done in characterizing the appearance of specular reflections. In this dissertation, we propose appearance representations based on a small number of photometric images for modeling specular and diffuse components. Two parts of this work utilize physical models of reflection, one for deriving a logarithmic specular representation using an approximated Torrance-Sparrow model, and one that involves estimation of material properties and extraction of implicit shape information for synthesis of diffuse and specular appearance. The third part of this work employs an empirical, but physically plausible, reflectance model for estimation of reflectance parameters. From this model, a technique for separation of reflection components and a linear representation for each reflection element are presented for enumerating the possible illumination appearances of an object. Potential applications of these photometric appearance models include object recognition and computer graphics rendering.