It's quite common to see web pages that show a solid color in place of images, while resources are being loaded. The color is usually averaged from the original image, as to give it a more "consistent" feeling while the page loads.

A simple way to do this from the command line is by using ImageMagick to resize the image to 1x1 size, then get the color of that one pixel:

% convert pic.jpg -resize '1x1!' txt:-
# ImageMagick pixel enumeration: 1,1,65535,srgb
0,0: (16818.3,27417.8,39342.8)  #416B99  srgb(65,107,153)

(The image being used in the examples is the photo at the beginning of this article, from Wikimedia).

The color looks right:

Now, onto how to get the same with Python.

A simple script like this would do:

from io import BytesIO

import requests
from PIL import Image

image_url = ''  # noqa

# Here we get the image from the web, but we could as easily load
# from the local filesystem. See Pillow docs on how to open images.
resp = requests.get(image_url)
assert resp.ok
img =

img2 = img.resize((1, 1))

color = img2.getpixel((0, 0))

This depends on the requests and pillow packages, make sure you have them installed:

% pip install requests pillow

The resulting hex color is different from the one from ImageMagick (although it looks "right" anyways)..

The reason is, ImageMagick uses antialias by default when resizing an image, thus the final color will be slightly different. If we do the same with Python:

img2 = img.resize((1, 1), Image.ANTIALIAS)

We will get the exact same color as with ImageMagick, #416b99.

Now, is it worth it? When I tried on a sample image, calculating antialiasing increased computation time by approx two orders of magnitude. Since we only care about the final "average" color, and both results look about right, I think we can stick with the quickest way.


Article cover photo: Arcalino / Wikimedia Commons / CC BY-SA 3.0