Display image as grayscale

python

image-processing

matplotlib

opencv

byNikita Barsukov·Feb 5, 2025

To display an image in grayscale in Python, use matplotlib's imshow function. This code defines the steps involved in reading, converting, and displaying this grayscale image.

from matplotlib import pyplot as plt
import matplotlib.image as mpimg

# Read the image (this assumes that the image was saved in the working directory)
img = mpimg.imread('your_image.png')

# Display the grayscale image
plt.imshow(img, cmap='gray', vmin=0, vmax=255)
plt.axis('off')  # This command is how we "tell" the axis "you have no power here!"
plt.show()

It's crucial to note that this code is assuming that your image has been converted to grayscale already.

Dissecting the grayscale pipeline

When examining grayscale display, understanding how these key components render color images into varying graytones, lies at the heart of it.

The RGB to grayscale journey

Understanding how our images turn into 50 shades of grey is quite simple. Color images are an RGB sandwich with Red, Green, Blue layers melding together to show us the colorful world. To convert it to grayscale, the sandwich is squished into one layer showing varying intensities of grey.

Pillow: The feather in your conversion cap

The Pillow library allows you to convert a color image to grayscale. Here's how:

from PIL import Image

# Open an image file
img_color = Image.open('your_image.jpg')
# Convert the image to grayscale
img_gray = img_color.convert('L')

With convert('L') we're changing our vibrant, bold, color image into a refined, sophisticated grayscale subject. Remember, L just means Luminance, a fancy term for brightness in an image.

Mastering grayscale display with matplotlib

Proper normalization of your grayscale image could be the difference between a moonlit night and a sunny day. Here's a foolproof plan:

import numpy as np
from matplotlib.colors import NoNorm

# Let the randomness commence! 
grayscale_matrix = np.random.randint(0, 50, (100, 100))

# Shhh... NoNorm is at work.
# We're telling the function to display without scaling the intensity 
plt.imshow(grayscale_matrix, cmap='gray', norm=NoNorm())
plt.axis('off')  # Do not disturb sign for our axes
plt.show()

For the vampire in you: Inverted grayscale

If you prefer your nights white and days black, simply use cmap='gray_r':

plt.imshow(img_gray, cmap='gray_r')
plt.axis('off') # Bye, bye axes!
plt.show()

Other options and potential hiccups

OpenCV way of life

OpenCV got your back for grayscale conversion and display:

import cv2

# Load an image in grayscale mode
img_gray_cv = cv2.imread('your_image.png', cv2.IMREAD_GRAYSCALE)

# Display grayscale glory with OpenCV
cv2.imshow('Grayscale Image', img_gray_cv)
cv2.waitKey(0) # Wait for the user to close the window 
cv2.destroyAllWindows() # We're environmentalists, we clean up after ourselves!

Troubleshooting the grayscale

Two main culprits can distort the grayscale display:

Normalization maladies: A sneak thief that distorts your image when you're not looking, especially for images with selective intensity ranges.
Colormap misuse: Choosing anything but gray or gray_r is like crashing a black and white party in a polka dot suit.

Customizing colormap range

Grayscale need not be boring. Let's make it interesting by playing with value range:

# Adjust vmin and vmax to flood some light into the grayscale
plt.imshow(img, cmap='gray', vmin=50, vmax=200)
plt.axis('off') # What's our mantra? Right. Axes off!
plt.show()