Page principale | Projets | Publications | Thèses | Présentations | Membres | Cours | Gallerie | Intérêt local Home | Projects | Publications | Theses | Presentations | People | Courses | Gallery | Local interest

Abstract

For many years now, simple local illumination models as introduced by Phong [Phong75], Blinn [Blinn77] and Cook [Cook82] have been successfully used to render 3D scenes with a certain degree of realism. Improvements for these basic models have been proposed with limited impact in the 3D graphics community.

This paper analyzes some of these proposed improvements and tries to determine why they are not widely used today. With some of these answers in mind, we establish some criteria that should be considered while developing a new illumination model and suggest possible improvements with respect to our criteria.

BibTeX entry

@InProceedings{Poulin:1992:SFL,
  author =       "Pierre Poulin",
  title =        "Separate functions for local illumination",
  pages =        "37--43",
  booktitle =    "Graphics Interface '92 Workshop on Local
                 Illumination",
  month =        may,
  year =         "1992",
  conference =   "held in Vancouver, B.C.; 11 May 1992",
  keywords =     "brdf, image synthesis, shading, illumination models,
                 realism",
  annote =       "Poulin proposes some reasons why reflection models
                 like Phong's, Blinn's and Cook's are still the most
                 commonly used today even though many other improvements
                 have been introduced. Criteria are suggested to follow
                 before introducing new improvements and the use of
                 separate functions is advocated to capture complexity
                 while remaining simple to define.",
}