This is a MATLAB implementation of the algorithm that solves a convex relaxation of Potts model discussed in "A Convex Relaxation Approach for Computing Minimal Partitions" by Pock et al. It is adapted from the code at https://github.com/VLOGroup/pgmo-lecture.
If you're using this code for your research, please cite Pock et al.'s paper.
@inproceedings{pock2009convex,
title={A convex relaxation approach for computing minimal partitions},
author={Pock, Thomas and Chambolle, Antonin and Cremers, Daniel and Bischof, Horst},
booktitle={2009 IEEE Conference on Computer Vision and Pattern Recognition},
pages={810--817},
year={2009},
organization={IEEE}
}
To run a demo, run one of the following scripts:
- script1.m
- script2.m
- script3.m
The general function to apply Potts model to an image is main_pott.m, which asks for the following arguments:
-f: image
-K: number of partitions/clusters
-lambda: fidelity parameter
-maxit: maximum number of iterations