implementation of the article "https://ieeexplore.ieee.org/document/5995323" . The code is implemented with the framework Networkx, and thus, it is very slow (as Networkx is pure python), in future version, I will replace Networkx with graph_tool or igrah (for which the main code is in C++), which is faster.
The syntax to call the code is the following: python main.py -img path_image -sp N where: path_image is the path to the image to be partitionned N is the number of superpixels for the partitionning of the image
choose a small image for tests