A number of CS students have been refactoring particular ImageMagick algorithms to take advantage of OpenMP. The greatest benefit of OpenMP would come from adapting the pixel cache (magick/cache.c) since this would speed up every algorithm (all algorithms pull and push pixels from/to the cache). In addition, the slowest ImageMagick algorithm is FxImage (magick/fx.c) since it is an interpreted pixel expression language. Adapting FxImage() to OpenMP would be of great benefit to the ImageMagick user community as well. Or even better, convert the interpreter to a postfix stack and evaluate the stack (the algortihm currently uses recursion which is slow).
If you choose to accept this challenge, post your intentions here and happy coding. _______________________________________________ Magick-users mailing list [email protected] http://studio.imagemagick.org/mailman/listinfo/magick-users
