Analytic Rasterization on GPGPUs

Information

• Publication Type: Poster
• Workgroup(s)/Project(s): not specified
• Date: July 2013
• Note: Best Poster Award
• Location: Barcelona
• Event: PUMPS Summer School
• Conference date: 8. July 2013 – 12. July 2013
• Keywords: hidden surface elimination, vector format, analytic, rasterization, CUDA, GPGPU, prefiltering, anti-aliasing, parallel, aliasing, closed-form, visibility, convolution, filter

Abstract

In this poster we present an overview of our work on analytic anti-aliased rasterization. In contrast to the traditional supersampling approach, we use exact prefiltering to remove frequencies above the Nyquist limit. This is enabled by performing exact hidden surface elimination to compute the visible regions of all scene primitives. We derived a closed-form solution to subsequent prefilter convolution which guarantees high-quality anti-aliasing up to numerical precision. Our analytic rasterization pipeline is based on several highly parallel algorithms and an efficient implementation in CUDA is presented as well as several result images of its performance on complex scenes.

Additional Files and Images

Weblinks

No further information available.

BibTeX

@misc{Auzinger_2013_AnaRaster,
  title =      "Analytic Rasterization on GPGPUs",
  author =     "Thomas Auzinger",
  year =       "2013",
  abstract =   "In this poster we present an overview of our work on
               analytic anti-aliased rasterization. In contrast to the
               traditional supersampling approach, we use exact
               prefiltering to remove frequencies above the Nyquist limit.
               This is enabled by performing exact hidden surface
               elimination to compute the visible regions of all scene
               primitives. We derived a closed-form solution to subsequent
               prefilter convolution which guarantees high-quality
               anti-aliasing up to numerical precision. Our analytic
               rasterization pipeline is based on several highly parallel
               algorithms and an efficient implementation in CUDA is
               presented as well as several result images of its
               performance on complex scenes. ",
  month =      jul,
  note =       "Best Poster Award",
  location =   "Barcelona",
  event =      "PUMPS Summer School",
  Conference date = "Poster presented at PUMPS Summer School
               (2013-07-08--2013-07-12)",
  keywords =   "hidden surface elimination, vector format, analytic,
               rasterization, CUDA, GPGPU, prefiltering, anti-aliasing,
               parallel, aliasing, closed-form, visibility, convolution,
               filter",
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2013/Auzinger_2013_AnaRaster/",
}