Sampling and reconstructing point clouds

Information

• Publication Type: Invited Talk
• Workgroup(s)/Project(s):
  • Modeling the World at Scale
• Date: 28. May 2024
• Event: TU Graz
• Location: Graz
• Keywords: curve reconstruction, surface reconstruction, sampling

Abstract

Curve and surface reconstruction from unstructured points represent a fundamental problem in computer graphics and computer vision, with many applications. The quest for better solutions for this ill-posed problem is riddled with various kinds of artifacts such as noise, outliers, and missing data.

Moreover, the reconstruction problem usually implies further input requirements: how many samples do we need for a successful reconstruction, what properties should these samples satisfy and how can we obtain such sets. And once we obtain these point samples, how can we extract connectivity that best approximates the initial surface they have been sampled from?

We will discuss about various sampling strategies, corresponding reconstruction methods, with multiple applications in automating sketch coloring, adaptive meshing for faster simulations, and cultural heritage.

Additional Files and Images

Weblinks

No further information available.

BibTeX

@talk{ohrhallinger_stefan-2024-inv,
  title =      "Sampling and reconstructing point clouds",
  author =     "Stefan Ohrhallinger and Diana Marin",
  year =       "2024",
  abstract =   "Curve and surface reconstruction from unstructured points
               represent a fundamental problem in computer graphics and
               computer vision, with many applications. The quest for
               better solutions for this ill-posed problem is riddled with
               various kinds of artifacts such as noise, outliers, and
               missing data.  Moreover, the reconstruction problem usually
               implies further input requirements: how many samples do we
               need for a successful reconstruction, what properties should
               these samples satisfy and how can we obtain such sets. And
               once we obtain these point samples, how can we extract
               connectivity that best approximates the initial surface they
               have been sampled from?  We will discuss about various
               sampling strategies, corresponding reconstruction methods,
               with multiple applications in automating sketch coloring,
               adaptive meshing for faster simulations, and cultural
               heritage. ",
  month =      may,
  event =      "TU Graz",
  location =   "Graz",
  keywords =   "curve reconstruction, surface reconstruction, sampling",
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2024/ohrhallinger_stefan-2024-inv/",
}