You can perform Bitwise operations at pixel level. Following bitwise operators are used in common.

a.   not

b.   and

c.    or

d.   xor

not

It inverts a bit value. If the bit has value 0, then it inverts to 1. If the bit has value 1, then it inverts to 0.

Let me use a binary image to demonstrate the examples.

image = np.zeros((500, 500), dtype='uint8')

cv.rectangle(image, (50, 50), (400, 400), color=255, thickness=-1)

Above snippet create a blank image and draw a rectangle on it. It looks like below.

When we apply not operation on the above image, all the black pixels become white and white pixels become balck.

image_with_not_operator = cv.bitwise_not(image)

Above snippet create below image.

Find the below working application.

not_operator.py

import cv2 as cv
import numpy as np

image = np.zeros((500, 500), dtype='uint8')
cv.rectangle(image, (50, 50), (400, 400), color=255, thickness=-1)

image_with_not_operator = cv.bitwise_not(image)

cv.imshow('image', image)
cv.imshow('image_with_not_operator', image_with_not_operator)

cv.waitKey(0)

# Close the OpenCV windows
cv.destroyAllWindows()

Bitwise or operator

It is used to combine or overlay two images. Following table summarizes how bitwise operator works.

Suppose we have two images like below.

When you apply bitwise or operator on both the images, you will get below image.

or_operator.py

import cv2 as cv
import numpy as np

image = np.zeros((500, 500), dtype='uint8')

blank_image_1 = image.copy()
blank_image_2 = image.copy()

cv.rectangle(blank_image_1, (40, 40), (450, 450), color=255, thickness=-1)
cv.circle(blank_image_2, (250, 250), 240, color=255, thickness=-1)

image_with_or_operator = cv.bitwise_or(blank_image_1, blank_image_2)

cv.imshow('Rectangle', blank_image_1)
cv.imshow('Circle', blank_image_2)

cv.imshow('image_with_or_operator', image_with_or_operator)

cv.waitKey(0)

# Close the OpenCV windows
cv.destroyAllWindows()

Bitwise and operator

It is used for masking. You can create a binary mask with certain pixel values set to 1 (white) and others to 0 (black). By applying a bitwise AND operation with the masked image, you can extract specific regions of an image.

Bitwise and operator generate below image.

and_operator.py

import cv2 as cv
import numpy as np

image = np.zeros((500, 500), dtype='uint8')

blank_image_1 = image.copy()
blank_image_2 = image.copy()

cv.rectangle(blank_image_1, (40, 40), (450, 450), color=255, thickness=-1)
cv.circle(blank_image_2, (250, 250), 240, color=255, thickness=-1)

image_with_and_operator = cv.bitwise_and(blank_image_1, blank_image_2)

cv.imshow('Rectangle', blank_image_1)
cv.imshow('Circle', blank_image_2)

cv.imshow('image_with_and_operator', image_with_and_operator)

cv.waitKey(0)

# Close the OpenCV windows
cv.destroyAllWindows()

Bitwise xor operator

When you XOR an image with a specific pattern, it can create visual effects or hide information within the image