Automatic Gradient-Preserving Stencilization of Raster Images

Information

  • Publication Type: Bachelor Thesis
  • Workgroup(s)/Project(s): not specified
  • Date: March 2021
  • Date (to): 2021
  • Matrikelnummer: 01527584
  • First Supervisor: Hsiang-Yun Wu

Abstract

Stencils are used as intermediate objects with designed gaps in them, to create patterns on surfaces by applying pigments on the surface through the stencil, which allows the pigment to reach the surface through the gaps and thereby to create the pattern on the surface. For the production of a stencil out of any raster image, it is not enough to assume the background color as the parts of the material that will be cut out and the other color as the material remaining in the stencil. There has to be cohesion between all the independent parts that are left in so that they do not have to be held in place individually. The needed connections between the components could be made very obvious and easy to distinguish from the intended shapes in order to draw over them later on with a paintbrush. The goal of this work however, will be an algorithm that produces connections between the shapes that can be left in the image the stencil produces, without disturbing the appearance of the shapes present (too much). This is done by finding the directions of the shapes’ contours on a vectorized version of the original image, to be able to continue in the same direction with the connections between different shapes. Then from all the possible connections the ones that will be used are found by creating a graph data structure and finding a maximum matching of that graph. In the end, it will be possible to input a binary image and get back a continuous stencil form that can be used as-is.

Additional Files and Images

Additional images and videos

Additional files

Weblinks

No further information available.

BibTeX

@bachelorsthesis{Kompatscher-2021,
  title =      "Automatic Gradient-Preserving Stencilization of Raster
               Images",
  author =     "Jan Kompatscher",
  year =       "2021",
  abstract =   "Stencils are used as intermediate objects with designed gaps
               in them, to create patterns on surfaces by applying pigments
               on the surface through the stencil, which allows the pigment
               to reach the surface through the gaps and thereby to create
               the pattern on the surface. For the production of a stencil
               out of any raster image, it is not enough to assume the
               background color as the parts of the material that will be
               cut out and the other color as the material remaining in the
               stencil. There has to be cohesion between all the
               independent parts that are left in so that they do not have
               to be held in place individually. The needed connections
               between the components could be made very obvious and easy
               to distinguish from the intended shapes in order to draw
               over them later on with a paintbrush. The goal of this work
               however, will be an algorithm that produces connections
               between the shapes that can be left in the image the stencil
               produces, without disturbing the appearance of the shapes
               present (too much). This is done by finding the directions
               of the shapes’ contours on a vectorized version of the
               original image, to be able to continue in the same direction
               with the connections between different shapes. Then from all
               the possible connections the ones that will be used are
               found by creating a graph data structure and finding a
               maximum matching of that graph. In the end, it will be
               possible to input a binary image and get back a continuous
               stencil form that can be used as-is.",
  month =      mar,
  address =    "Favoritenstrasse 9-11/E193-02, A-1040 Vienna, Austria",
  school =     "Research Unit of Computer Graphics, Institute of Visual
               Computing and Human-Centered Technology, Faculty of
               Informatics, TU Wien ",
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2021/Kompatscher-2021/",
}