Skip to content

roflmaostc/Fast-Approximate-Energy-Minimization-via-Graph-Cuts

Repository files navigation

Fast-Approximate-Energy-Minimization-via-Graph-Cuts

This repository contains stuff about Fast Approximate Energy Minimization via Graph Cuts. In the figures folder you will find many figures that you can use. They are all made witz tikz and LaTeX. The report_fcv.tex and paper_report.tex won't compile since a few figures and the template are missing.

The code implements the swap-algorithm according to the paper Fast Approximate Energy Minimization via Graph Cuts from Boykov, Veksler and Zabih. The code is written in Python 3. For the Graph-Cut we use PyMaxflow. Install requirements:

sudo pip install -r requirements.txt

Executing is done by:

python minimization.py path/filename.png cycles

Cycles can be an arbitrary number. Something below 10 ist quite good. Hint: Don't use images larger than 100x100 pixel, it will take a lot of time otherwise.

About

Minimize Energy in Images.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published