| OCR Text |
Show CHAPTER 1 INTRODUCTION As part of the field of realistic image synthesis, shading models for computer graphics are concerned with converting geometric representations of objects to images. The goal is to generate images that correspond closely to the way a model would appear if it were a genuine physical object under some actual lighting conditions. Given a geometric model and its material characteristics, many issues are involved in image generation and display. Illumination, reflection, transparency, color reproduction, and antialiasing for raster display are all part of the process. Shading models specify what information will be used to model illumination and reflection and how that information will be used. With current methods, a shading model is evaluated one or more times per displayed pixel. At common display resolutions, several hundred thousand pixels are in an image, so computational costs of a shading model are a serious consideration. However, in realistic image synthesis the goal is to fool the eye, which is not a simple task. Sophisticated illumination models are necessary to creating convincing images, so efficiency and accuracy in shading models are topics of continuing research. In the past four years, interest has been stirred in the computer graphlcs community regarding the role of diffuse reflection for shading. Early shading methods did not account for indirect lighting and color bleeding effects. This thesis follows other research in using radiosity methods derived from radiative heat transfer to treat diffuse reflection. |
