Pano.NET - An interactive application for camera calibration, image stitching and projective operations

Information

• Publication Type: Bachelor Thesis
• Workgroup(s)/Project(s):
  • Photo-Guide: Image-Based City Exploration
• Date: September 2013
• Date (Start): 3. April 2013
• Matrikelnummer: 0927289
• First Supervisor:
  • Michael Birsak
  • Przemyslaw Musialski
• Keywords: multi-view image processing, panorama imaging, image compositing

Abstract

The automatic generation of panoramic image mosaics has evolved into a mainstream consumer application in recent years, thanks to modern algorithms and methods that can generate beautiful- looking, artifact-free panoramas with minimal user interaction. Simple panoramas are only one possible application, other examples such as the scene-aware Google Street View show what is possible using these technologies. Software libraries such as OpenCV allow integration of image stitching and related operations into a multitude of applications with relative ease. In this thesis, we present an interactive application (Pano.NET) that allows the user to experiment with several commonly used methods for camera calibration and image stitching (as implemented in OpenCV), and export related data such as registration information. We also give an overview of the used algorithms. In addition, we implemented a projective drawing system that allows the user to draw lines and polygons onto a panorama and on the views that it consists of. Finally, a method is presented that can generate six-sided cube maps out of a spherical panorama.

Additional Files and Images

Weblinks

No further information available.

BibTeX

@bachelorsthesis{schaukowitsch-2013-fls,
  title =      "Pano.NET - An interactive application for camera
               calibration, image stitching and projective operations",
  author =     "Florian Schaukowitsch",
  year =       "2013",
  abstract =   "The automatic generation of panoramic image mosaics has
               evolved into a mainstream consumer application in recent
               years, thanks to modern algorithms and methods that can
               generate beautiful- looking, artifact-free panoramas with
               minimal user interaction. Simple panoramas are only one
               possible application, other examples such as the scene-aware
               Google Street View show what is possible using these
               technologies. Software libraries such as OpenCV allow
               integration of image stitching and related operations into a
               multitude of applications with relative ease. In this
               thesis, we present an interactive application (Pano.NET)
               that allows the user to experiment with several commonly
               used methods for camera calibration and image stitching (as
               implemented in OpenCV), and export related data such as
               registration information. We also give an overview of the
               used algorithms. In addition, we implemented a projective
               drawing system that allows the user to draw lines and
               polygons onto a panorama and on the views that it consists
               of. Finally, a method is presented that can generate
               six-sided cube maps out of a spherical panorama.",
  month =      sep,
  address =    "Favoritenstrasse 9-11/E193-02, A-1040 Vienna, Austria",
  school =     "Institute of Computer Graphics and Algorithms, Vienna
               University of Technology ",
  keywords =   "multi-view image processing, panorama imaging, image
               compositing",
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2013/schaukowitsch-2013-fls/",
}