Evolutionary Image Segmentation Based on Multiobjective Clustering
This is a Python implementation of the following paper:
If you use this code for your research, please cite our paper:
@inproceedings{ShirakawaCEC2009,
author = {Shinichi Shirakawa and Tomoharu Nagao},
title = {Evolutionary Image Segmentation Based on Multiobjective Clustering},
booktitle = {Proceedings of the 2009 IEEE Congress on Evolutionary Computation (CEC 2009)},
pages = {2466--2473},
year = {2009}
}
Requirements
We tested the codes on the following environment:
- Python 3.6.0
- Python package version
- NumPy 1.16.2
- SciPy 0.19.1
- Matplotlib 2.0.2
- cv2 3.4.0
- Numba 0.35.0
- DEAP 1.3
Usage
- Run the python script as
python mock_segmentation.py
- In the default setting, the program loads
paprika.png
as the input image and uses RGB color space - After execution, the result (output images and a graph) is saved in
./out/
- If you want to use another image file, please add
-i
option aspython mock_segmentation.py -i your_image.png
- If you want to use Lab* color space, please add
-c
option aspython mock_segmentation.py -c Lab
- If you want to run the code with a different setting, please directly modify the script (parameters are set in the beginning of
mock_segmentation.py
)