The scripts used in Color Bars take the movie posters of IMDb's top movies and performs k-means clustering to produce a color quantized version of each poster. Then, each color quantized poster is used to produce a color bar representing the proportions of colors in the poster.
$python color_clustering.py --help
usage: color_clustering.py [-h] -i IMG -k CLUSTERS
optional arguments:
-h, --help show this help message and exit
-i IMG, --img IMG Image path
-k CLUSETERS, --clusters CLUSTERS
Number of clusters
$python imdb_posters.py --help
usage: imdb_posters.py [-h] -k CLUSTERS
optional arguments:
-h, --help show this help message and exit
-k CLUSTERS, --clusters CLUSTERS
Number of clusters
To view all 250 color bars, see color_bars_4 for k = 4 and color_bars_8 for k = 8.