Friedrich-Alexander-Universität DruckenUnivisEnglish FAU-Logo
Techn. Fakultät Willkommen am Department Informatik FAU-Logo
Codesign
Lehrstuhl für Informatik 12
CUDA
Department Informatik  >  Informatik 12  >  Personal  >  Richard Membarth  >  CUDA

CUDA Challenge:

The following GIMP plugin was awarded with the first prize of the CUDA Challenge (official announcement, German version). The plugin applies the multiresolution filter investigated in my diploma thesis to grayscale images. The filter smooths an image while it preserves the details of interest. The filter was implemented using CUDA and runs on the graphics card. Compared to a fully parallelized CPU version on a Xeon Quad Core a speedup of more than 60 can be achieved on high end graphics card like the Tesla C1060. Compared to a single core implementation, the speedup is even higher. As seen in the screenshot below, the size of the filter can be adjusted (geometric spread) and also a threshold value, which determines up to which level details of the images get smoothed (photometric spread):

gimp_plugin.png
The source code of the plugin can be downloaded here and requires CUDA 3.0 or above (also compatible with CUDA 2.0 and above), the CUDA SDK, as well as GIMP and its development libraries. The plugin was successfully run and tested on various GNU/Linux systems and on Mac OS X.
Note: For Mac OS X Snow Leopard, a 64-bit GIMP installation, for example from MacPorts, was used.
  Impressum Stand: 24 August 2010.   R.M.