Do we need the full reconstruction pipeline?

Information

• Publication Type: Invited Talk
• Workgroup(s)/Project(s):
  • Harvest4D - Harvesting Dynamic 3D Worlds from Commodity Sensor Clouds
• Date: 2014
• Event: EU-Korea Conference on Science and Technology
• Location: Vienna, Austria
• Conference date: 24. July 2014

Abstract

The traditional cultural heritage documentation pipeline from acquisition using a range scanner to interactive display to the user is a tedious and labor-intensive process. In particular, reconstructing high-quality meshes from large point clouds can be time consuming. In this talk, I will present shortcuts to this pipeline. The first idea is not to reconstruct a mesh at all, but keep the original point cloud as long as possible. I will discuss the challenges in maintaining interactivity and high quality when dealing with the display and manipulation of huge point clouds. The second idea is to reconstruct extremely simple models for regular and man-made structures, using shape analysis and user guidance. These models can be shown in end-user installations and require very few resources for display.

Additional Files and Images

Weblinks

No further information available.

BibTeX

@talk{WIMMER-2014-DWNT,
  title =      "Do we need the full reconstruction pipeline?",
  author =     "Michael Wimmer",
  year =       "2014",
  abstract =   "The traditional cultural heritage documentation pipeline
               from acquisition using a range scanner to interactive
               display to the user is a tedious and labor-intensive
               process. In particular, reconstructing high-quality meshes
               from large point clouds can be time consuming. In this talk,
               I will present shortcuts to this pipeline. The first idea is
               not to reconstruct a mesh at all, but keep the original
               point cloud as long as possible. I will discuss the
               challenges in maintaining interactivity and high quality
               when dealing with the display and manipulation of huge point
               clouds. The second idea is to reconstruct extremely simple
               models for regular and man-made structures, using shape
               analysis and user guidance. These models can be shown in
               end-user installations and require very few resources for
               display. ",
  event =      "EU-Korea Conference on Science and Technology",
  location =   "Vienna, Austria",
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2014/WIMMER-2014-DWNT/",
}