Turning images into '9-pan' palettes using KMeans clustering from sklearn.
We require:
- Pillow, for opening and processing images
- Scikit Learn, for clustering
We use numpy. Since it's a dependency of scikit-learn, we're not specifying it; we're going to use the version that comes with our pinned sklearn version.
On Raspberry Pi, we ran into the error
Original error was: libf77blas.so.3: cannot open shared object file: No such file or directory
So we did the following:
sudo apt-get install libatlas-base-dev
The numpy developer documentation recommended either doing that or installing the version of numpy packaged for raspbian. Since we want to use the version of numpy included with sklearn for the least number of dependency headaches, we install libatlas instead.
On Linux Mint, I found I needed to install imagemagick for the 'display' command. apt search imagemagick for additional details.
If you run into additional issues running the script, please add an Issue with your problem or solution to this repository. If you don't have a solution, I'll do my best to come up with one.
We recommend a virtual environment.
~$ python3 -m venv venv
~$ source venv/bin/activate
~$ python3 -m pip install -r requirements.txt
Once that process is complete, run the program:
python3 img2palette.py -i <your image>
To run in express mode, pass -x:
python3 img2palette.py -x -i <your image>
The output is OK. We should tweak the options in the future.
For this image by Marco Ferrarin:
We receive this palette:
Perhaps the clustering could be adjusted for different / better results? But, for a first attempt, I'm pretty happy.