An implementation of the paper Depixelizing Pixel Art by Kopf and Lischinski. The implementation is purely in python and instructions are given below.

Instructions to run

The implementation can be run by using the command "python main.py" wherein, the command line asks you to enter the path of the input image.

The output image will be created with the same name as the input image in the "outputs/" folder (except with a .svg extension).

Depending on the input file's dimensions, the running time can be long.

There are sample testcases (around 25) in the "testcases/" folder.

You can view some of the outputs that we have got, in the "outputs/" folder.

Requirements

• Python 2.7
• pypng : https://pypi.python.org/pypi/pypng
• networkx: https://pypi.python.org/pypi/networkx/
• svgwrite : https://pypi.python.org/pypi/svgwrite/

Citations

• https://johanneskopf.de/publications/pixelart/paper/pixel.pdf
• https://github.com/gityou/depixelize
• http://dev.crazyrobot.net/2011/07/12/pixelart-to-vector/
• https://code.launchpad.net/libdepixelize
• http://networkx.lanl.gov/index.html
• https://github.com/jerith/depixel
• https://svgwrite.readthedocs.org/en/latest/
• http://vinipsmaker.wordpress.com/2013/07/21/splines-extraction-on-kopf-lischinski-algorithm-part-1/
• http://vinipsmaker.wordpress.com/2013/08/13/splines-extraction-on-kopf-lischinski-algorithm-part-2/

Authors

Anirudh Vemula, Vamsidhar Yeddu

Computer Science, IIT Bombay