Usage

modern_colorthief exposes two functions get_color and get_palette

Image Path

How to use with image path:

import modern_colorthief

# Path to any image
path = ...

print(modern_colorthief.get_palette(path)) # returns list[tuple[int,int,int]]
print(modern_colorthief.get_color(path)) # returns tuple[int,int,int]

BytesIO

How to use modern_colorthief with a BytesIO object:

import io

import modern_colorthief

path = ...

image_bytes = io.BytesIO()

dominant_color = modern_colorthief.get_color(image_bytes)
dominant_palette = modern_colorthief.get_palette(image_bytes)

Pillow Object

How to use modern_colorthief with a Pillow object:

import io
from PIL import Image

import modern_colorthief

path = ...
img = Image.open(path, mode="r")

image_bytes = io.BytesIO()
img.save(image_bytes, format="PNG")

dominant_color = modern_colorthief.get_color(image_bytes)
dominant_palette = modern_colorthief.get_palette(image_bytes)

Numpy Array

How to use modern_colorthief with numpy array (needs Pillow):

import numpy as np
from PIL import Image

arr = ... # Numpy Array

img = Image.fromarray(arr)

image_bytes = io.BytesIO()
img.save(image_bytes, format="PNG")

dominant_color = modern_colorthief.get_color(image_bytes)
dominant_palette = modern_colorthief.get_palette(image_bytes)

If there is a difference in the colors returned by modern_colorthief check the parity documentation