Perform color quantization by selecting the palette randomly or using K-means.
Setup the environment using conda:
conda env create -f environment.yml
Here's a sample usage that displays the output of the color quantization performed on the same image using random selection and K-means:
from image_quantizer import quantizer
q = quantizer.ImageQuantizer()
qimages = q.quantize_multi([
{'n_colors': 8, 'method': 'random'},
{'n_colors': 16, 'method': 'random'},
{'n_colors': 32, 'method': 'random'},
{'n_colors': 8, 'method': 'kmeans'},
{'n_colors': 16, 'method': 'kmeans'},
{'n_colors': 32, 'method': 'kmeans'}
], image_filename='tests/fixtures/Lenna.png')
quantizer.compare(*qimages)
The output will be something as follows: