Computer stuff

December 9th, 2008

Programming Ideas
Genetic Algorithms for Vectorization of a Bitmap
Mona Lisa
Academic paper

Other links
Reddit question
ECJ

For fun and learning, I would like to replicate what those gentlemen have done using JavaFX. What is weird is how the generated vector graphics don’t seem to follow the “natural” edges in the photos. In particular the, in the pdf link above, the results of the “Union Jack” approximation.

The twist I would like to add is to do an edge detection of the source photo and use that information in the genetic alorithm.

Two way I could see this being used.

  1. As a way to choose an initial population of chromosomes. Aka, make all the starting endpoints be on points from the edge detection run.
  2. Incorporate the edge detection into the fitness function. Such as rasterize the image as the other guys have done, but also rasterize the image just drawing the edges of the triangle, and compare that with the edge detection “image”.

The twist I would like to make to possibly address the “sharpness” of the vectorization is to include in the fitness function a

Linux

Ubuntu Performace Guides

Comments are closed.