Mathematical Operations On Images Using OpenCV and NumPy in Python

In this article, you will learn to:

• How to Add Two Images

• How to Subtract Two images

• How to Multiply Two Images

We first load and display two images. For the above operations, the images should be of the same size and type, the second image may be of scalar type.

Install and Import Necessary Libraries:

Install OpenCV using pip command.

pip install opencv-python

First, we import all necessary libraries like cv2, NumPy, and matplotlib, etc.

import cv2
import numpy as np
import matplotlib.pyplot as plt

Read/load and Display Images

import cv2

# load first image
img1 = cv2.imread('lena.png')
img1 = cv2.cvtColor(img1, cv2.COLOR_BGR2RGB)

# load second image 
img2 = cv2.imread('OpenCV_logo.png')
img2 = cv2.cvtColor(img2, cv2.COLOR_BGR2RGB)

# resize the images
img1 = cv2.resize(img1,(500,500))
img2 = cv2.resize(img2, (500,500))

# display first image
cv2.imshow('image',img1)
cv2.waitKey(0)
cv2.destroyAllWindows()

# display second image
cv2.imshow('image',img2)
cv2.waitKey(0)
cv2.destroyAllWindows()

Output:

   

You may be interested in the below post to learn different ways to load images.

Different Ways to Load Images in Python from Scratch,

How to Add Two Images

Using cv2.add() Method

We can use OpenCV function  cv2.add() to add two images, or simply we can use Numpy operation img1 + img2 to add the images. 

cv2.add()

img1 = cv2.imread('lena.png')
img2 = cv2.imread('OpenCV_logo_white.png')
img = cv2.add(img1, img2)  # 240+22 = 262 => 255

cv2.imshow('image',img)
cv2.waitKey(0)
cv2.destroyAllWindows()

Output:

Notice that cv2.add() method adds two images of same size, i.e., same height and width. So we have resized the images to make them of same size.

We can add or subtract images of different sizes using Python-OpenCV without resizing them.  

Using numpy operation  img1 + img2 

img1 = cv2.imread('lena.png')
img2 = cv2.imread('OpenCV_logo_white.png')
img = img1 + img2   # 240+22 = 262 % 256 => 6

cv2.imshow('image',img)
cv2.waitKey(0)
cv2.destroyAllWindows()

Output:

Note the difference between the above two resultant images.

Image Blending

If you want to add two images with different weights, use cv2.addWeighted() function

    img=α⋅img1+β⋅img2+γ

Here α and β are weights of first and second images and γ is a scalar quantity.

img1 = cv2.imread('lena.png')
img2 = cv2.imread('OpenCV_logo_white.png')
img = cv2.addWeighted(img1, 0.3, img2, 0.7, 0) 

cv2.imshow('image',img)
cv2.waitKey(0)
cv2.destroyAllWindows()

Output:

Add Two Images Vertically and Horizontally

We use cv2.vconcat([image1, image2]) and  cv2.hconcat([image1, image2]) to add two images vertically and horizontally respectively. 

import cv2
import matplotlib.pyplot as plt
#add images virtically
img_v = cv2.vconcat([img1,img2])
#add images horizontally
img_h = cv2.hconcat([img1,img2])
fig = plt.figure(figsize=(5, 5))
rows, columns = 1, 2
fig.add_subplot(rows, columns, 1)
# showing vertical image
plt.imshow(img_v)
plt.title("Virtical Add")
fig.add_subplot(rows, columns, 2) 
# showing horizontal image
plt.imshow(img_h)
plt.title("horizontal Add")

Output:

Vertically and Horizontally added image output

Note: Here, to display images we used plt.imshow() to show the pixel values.

We can also use np.vstack(), and np.hstack() to add two images vertically and horizontally.

import numpy as np
vertical = np.vstack((img1, img2))
horizontal = np.hstack((img1, img2))

How to Subtract Two Images

# subtract images
sub_cv = cv2.subtract(img1, img2)
sub_np = np.subtract(img1,img2)
sub_abs = cv2.absdiff(img1, img2)
# define figure to display above three images
fig = plt.figure(figsize=(5, 5))
rows, columns = 2, 2
# add first image to figure
fig.add_subplot(rows, columns, 1)
# showing image
plt.imshow(sub_cv)
plt.axis('off')
plt.title("cv2 image subtraction")
# add second image to figure
fig.add_subplot(rows, columns, 2) 
# showing image
plt.imshow(sub_np)
plt.axis('off')
plt.title("np image subtraction")
# add third image to figure
fig.add_subplot(rows, columns,   
# showing image
plt.imshow(sub_abs)
plt.axis('off')
plt.title("cv2 abs diff")

Output:

Image Subtraction 

How to Multiply Two Images:

#multiply images
mul_cv = cv2.multiply(img1, img2)
mul_np = np.multiply(img1, img2)
# sub_abs_np = np.abs(img1,img2)
fig = plt.figure(figsize=(5, 5))
rows, columns = 1, 2
fig.add_subplot(rows, columns, 1)
  
# showing image
plt.imshow(mul_cv)
plt.axis('off')
plt.title("cv2 image multiplication")
fig.add_subplot(rows, columns, 2)
  
# showing image
plt.imshow(mul_np)
plt.axis('off')
plt.title("np image multiplication")

Output:

Image Multiplication

Further Reading:

• How to Add/ Subtract different size images using OpenCV Python?
• How to Load, Pre-process and Visualize CIFAR-10 and CIFAR -100 datasets
• How to calculate mean and standard deviation of images in PyTorch
• Basic Operations on Images using OpenCV in Python
• Different ways to load images in Python
• How to normalize image dataset in PyTorch

Useful Resources:

• Computer Vision
• How To Posts

Next Post: How to Load and Visualize Multiple Images in Python?

Previous Post: How to compute mean, standard deviation, and variance of a tensor in PyTorch?