A Python script that takes an image and outputs a pixelated version using Voronoi tessellations.
The script loads the image, places n_points random points in the image and uses them as the seeds in a Voronoi tessellation. Then, region of the tessellation gets covered by the average color in it.
Right now the script is somewhat slow, because I went with the overkill of computing the shortest distance of each pixel to the seeds using a KDTree.
I assume you are running Python 3. You can create a new environment and install the requirements using
pip install -r requirements.txt
Then, you can run:
python process.py path_to_image.jpg
After running this script, the generated image will be saved as out.jpg in your working directory.
You can also specify two parameters: out (the path and name of the resulting image) and n_points (the amount of random points that are used as seeds in the tessellation). For example
python process.py example.jpg --out="example_50.jpg" --n_points=50
