The Coding Couple

Pair programming is a lifetime commitment.

Counting Pixels by Color in Python with Pillow (a PIL fork)

by

I’m a maker and love creating 8-bit art, mostly with wood. I have a new project in the works where I need an exact count of each pixel per a color. I started to manually count them before I stopped myself and thought, why don’t I just write a script to automate this!? Thus pixel-color-count.py was born!

Pixel Color Count

Given a valid image file, the Python script will iterate through each pixel in an image keeping a running tally of how many times the color of the pixel has appeared in the image.

Once the loop is done, the script will print to the console a list of each color and the number of times the color was present in the image.

Sample Output of pixel-color-count.py

Sample out generated from running the Python script

Built With

Pillow

I used Pillow, a fork of PIL (Python Image Library), for the image manipulation. This is my second project using Pillow. It’s a really great library for programmatically editing or manipulating images. 

with Image.open(args.image) as image:
        color_count = {}
        width, height = image.size
        rgb_image = image.convert('RGB')

The main functionality from the Image module I’m using is the getpixel method. 

rgb = rgb_image.getpixel((x, y))

webcolors

A second dependency for the project is more of a nice to have: webcolors. Given a color in RGB format, webcolors will return a human readable name for it if applicable. 

name = webcolors.rgb_to_name(rgb)

webcolors.rgb_to_name will throw a ValueError is it cannot find a name for a RGB color. In my script, I fall back to displaying the color as a tuple.

What’s left to implement?

In my first iteration of the script, I started to implement an optional command line argument for ignored colors. The idea behind this is there may be some colors I don’t care to count therefore I wouldn’t want my output cluttered.

parser.add_argument('-i', '--ignore-color', type=tuple, help='Skip counting pixels of this color')

If the time permits, I’d love to go back and implement this feature. I hit a snag when I tried to use tuples as an argument type with argparse.

UPDATE #1: The --ignore-color command line option is partially implemented, but there’s no input validation. I basically accept the input as a string and use ast.literal_eval to convert it to a tuple.

Version 2 of the script, I plan on outputting a new version of the original image with a pixel legend and count per unique pixel color.

UPDATE #2: A version 2 of the script now exists that outputs an image featuring a legend with a color square matched with the color name and pixel count.

Legend Image sample output for the Pixel Color Count script

And, maybe for a version 3 turn it into a web accessible app.

UPDATE #3: A web version of the pixel color counter now exists. View the source for the project here.

For now, the script does just what I need and has saved me a lot of effort over counting the manual way!

pixel-color-count.py

Check the GitHub repo for new updates to the script. Here’s there first iteration:

# pixel-color-count.py – Gets a sum of pixels per unique color
# maintainer – Ashley Grenon @townsean
import argparse
try:
from PIL import Image
except ImportError:
exit("This script requires the PIL module. Install with pip install Pillow")
try:
import webcolors
except ImportError:
# https://pypi.org/project/webcolors/
exit("This script uses webcolors for displaying named colors. Install with pip install webcolors")
def main():
parser = argparse.ArgumentParser(description='Calculates the sum of pixels per a color')
parser.add_argument('image', nargs='?', default='.', help='The image to sum the pixels per a color of')
# parser.add_argument('-i', '–ignore-color', type=tuple, help='Skip counting pixels of this color')
args = parser.parse_args()
with Image.open(args.image) as image:
color_count = {}
width, height = image.size
rgb_image = image.convert('RGB')
# iterate through each pixel in the image and keep a count per unique color
for x in range(width):
for y in range(height):
rgb = rgb_image.getpixel((x, y))
if rgb in color_count:
color_count[rgb] += 1
else:
color_count[rgb] = 1
print('Pixel Count per Unique Color:')
print('-' * 30)
color_index = 1
for color, count in color_count.items():
# if color == args.ignore-color:
# pass
try:
print('{}.) {}: {}'.format(color_index, webcolors.rgb_to_name(color), count))
except ValueError:
print('{}.) {}: {}'.format(color_index, color, count))
color_index += 1
if __name__ == '__main__':
main()
view raw pixel-color-count.py hosted with ❤ by GitHub

Find us online!