How to apply image thresholding in Python with NumPy

Learn how to apply a threshold to images in Python with numpy and Pillow.

Learn how to apply a threshold to images in Python with numpy and Pillow.

This snippet loads an image, applies a threshold, and then saves a new binarized image with only black and white pixels.

#more

import numpy as np
from PIL import Image


threshold = 100

img = Image.open('img_gray.png') 
# Note, if you load a color image, also apply img.convert('L')

img_np = np.array(img)
img_np = np.where(img_np > threshold, 255, 0)

img.putdata(img_np.flatten())

img.save('img_thresholded.png')