Color Distribution Transfer For Mixed-Reality Applications

Information

• Publication Type: Bachelor Thesis
• Workgroup(s)/Project(s):
  • Reciprocal Shading for Mixed Reality
• Date: October 2012
• Date (Start): 1. March 2012
• Date (End): 10. October 2012
• Matrikelnummer: 0925601
• First Supervisor:
  • Michael Wimmer
  • Martin Knecht
• Keywords: Augmented Reality, Color Transfer

Abstract

In mixed-reality environments it is essential to integrate virtual objects seamlessly into a real scene. Virtual objects should have similar appearances to those of real objects captured by a video camera. This is useful for many real-life application scenarios, including product advertising and visualization, edutainment systems or for enhancing cultural heritage sites.

Typical problems in this domain are to match the current ‘color mood’ of the video camera scene with the colors of virtual (rendered) objects. The color mood depends on the global illumination conditions as well as the hue, saturation or white balance settings of the camera.

The aim of this paper is to integrate existing methods of histogram transfers used in the domain of computational photography into mixed-reality environments. These methods allow us to simulate current luminance conditions in the scene and changes in the camera driver settings to apply them onto virtual objects.

This thesis contains two fast-running approaches to provide a color mapping between virtual objects and the real scene, which can be used in real-time applications. The results show that these methods increase the immersion of virtual objects in a real scene.

Additional Files and Images

Weblinks

No further information available.

BibTeX

@bachelorsthesis{spelitz_stefan-2012-CDTFMR,
  title =      "Color Distribution Transfer For Mixed-Reality Applications",
  author =     "Stefan Spelitz",
  year =       "2012",
  abstract =   "In mixed-reality environments it is essential to integrate
               virtual objects seamlessly into a real scene. Virtual
               objects should have similar appearances to those of real
               objects captured by a video camera. This is useful for many
               real-life application scenarios, including product
               advertising and visualization, edutainment systems or for
               enhancing cultural heritage sites.  Typical problems in this
               domain are to match the current ‘color mood’ of the
               video camera scene with the colors of virtual (rendered)
               objects. The color mood depends on the global illumination
               conditions as well as the hue, saturation or white balance
               settings of the camera.  The aim of this paper is to
               integrate existing methods of histogram transfers used in
               the domain of computational photography into mixed-reality
               environments. These methods allow us to simulate current
               luminance conditions in the scene and changes in the camera
               driver settings to apply them onto virtual objects.  This
               thesis contains two fast-running approaches to provide a
               color mapping between virtual objects and the real scene,
               which can be used in real-time applications. The results
               show that these methods increase the immersion of virtual
               objects in a real scene.",
  month =      oct,
  address =    "Favoritenstrasse 9-11/E193-02, A-1040 Vienna, Austria",
  school =     "Institute of Computer Graphics and Algorithms, Vienna
               University of Technology ",
  keywords =   "Augmented Reality, Color Transfer",
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2012/spelitz_stefan-2012-CDTFMR/",
}