| 摘要 |
A system and process for efficiently representing an object and allowing the synthesizing of photo-realistic images of the object that depict both diffuse and specular reflections therefrom. This is accomplished using a sparse set of input images and a geometric model of the object. In particular, the reflectance components (i.e., diffuse and specular) are separated out of the input images based on the intensity variation of object surface points. The diffuse reflection component is characterized with a global texture map, while the specular reflection components are used to model the illumination distribution of the environment surrounding the object and to estimate the surface reflectance parameters. This framework provides a very compact representation of the object's appearance. Photo-realistic virtual images of the object from any desired viewpoint can be synthesized using the previously computed global texture map, illumination distribution, and reflectance parameters.
|
