How to Add Two Images of Different Size in OpenCV Python?

Posted by on

In this post, we learn how to add two images of different sizes using OpenCV Python. We take two images of different sizes and perform pixel wise addition of these images. We use OpenCV Python method such as cv2.add(), cv2.addWeighted(), etc. to perform the operation.

Add images of different size in opencv python
Different Output Images after Adding Two Images

We will cover the contents listed in the following Table of Content.

Table of Content:

We will be using the following two images as the inputs to the programs discussed in this post.
Desert.jpg
Desert.jpg

Pneguins.jpg
Penguins.jpg

Prerequisites
  • Python 3
  • OpenCV
  • NumPy
  • Matplotlib (Optional)

Import Required Libraries/ Packages

The first step of any python programming is importing the required packages and libraries. We import the OpenCV (cv2) and Matplotlib. Using Matplotlib is optional. We used this to display the images with width and height. 
import cv2
import matplotlib.pyplot as plt

Load and Visualize the Images

The next step is to load the images. We load to images "Desert.jpg" and "Penguins.jpg". You can download the images and follow the process with the same images

Here you should change the image path accordingly.
# Load the input images 
img1 = cv2.imread('Desert.jpg')
img2 = cv2.imread('Penguins.jpg')
Now we should visualize the loaded image to make sure, we have loaded correct image. This part is optional.
# Visualize the first image
cv2.imshow('added image',img1)
cv2.waitKey(0)
cv2.destroyAllWindows()
Here we used OpenCV to display/ visualize the images.
To Visualize using Matplotlib follow the below two steps. Fist step is to convert images from BGR to RGB format and then visualize.
# Convert them from BGR to RGB
img1 = cv2.cvtColor(img1, cv2.COLOR_BGR2RGB)
img2 = cv2.cvtColor(img2, cv2.COLOR_BGR2RGB)
Now visualize first image- Desert.jpg
# Visualize First Image
plt.imshow(img1)
plt.show()
output image desert.jpg
Output- Desert.jpg

# Visualize Second Image
plt.imshow(img2)
plt.show()

Output image Penguins.jpg
Output -Penguins.jpg

We have discussed different ways to load image in Python in our previous post. 

Create ROI

The main task while adding or subtracting the images of different sizes is to find the ROI (Region Of Interest). Here our main aim is to find the minimum height and width from both the images. So that we can add the common area.  

Get height and width of images

First find the height and width of the images.
# Get height and width of first image
height1, width1 = img1.shape[0],img1.shape[1]
print(height1, width1)
Output
768  1024

# Get height and width of the second image
height2, width2 = img2.shape[0],img2.shape[1]
print(height2, width2)
Output
457  640

Notice the two images have different sizes (height and width).

Find minimum height and width of images

Now find the minimum height and width of the images.
# Find the minimum height and width of the two images
min_height = min(height1, height2)
min_width = min(width1, width2)
print(min_height, min_width)
Output

457  640

Crop images with minimum height and width

Crop both images with the above calculated minimum height and width. 
# Crop images with minimum height and width
img11 = img1[0:min_height, 0:min_width]
img22 = img2[0:min_height, 0:min_width]
Now visualize the cropped images
# Visualize first cropped image
plt.imshow(img11)
plt.show()
Output:
Cropped image 1
Cropped image (Desert.jpg) with min height and width

Visualize second cropped image.
plt.imshow(img22)
plt.show()
Output:
Cropped image (Penguins.jpg) with min height and width
Cropped image (Penguins.jpg) with min height and width

Now the both images have the same height and width. We could now add these images.

Note: You can OpenCV to visualize the output images. We are using the Matplotlib to show the scale for height and width. 

Create ROI by adding Cropped Images

Now add the cropped images with minimum height and width. This will be our ROI. We can use cv2.add() method to add the images of same size.
# Creatte ROI by adding the cropped images
img_add = cv2.add(img11, img22)
# Show the added miminum images
plt.imshow(img_add)
plt.show()
Output:
ROI-image of min height and width
ROI-image of min height and width

Alternatively you can use cv2.addWeighted() to add the images with different weights of images.
# Add minimum images with different weights
img_add = cv2.addWeighted(img11, 0.3, img22, 0.7, 0) 
plt.imshow(img_add)
plt.show()
We have added the common part of images. We can add this image part to the original images to find the final added images.

Add Images

Let's now move to the final task to add the images. We add the cropped ROI to the original images.

Finally add ROI to original Images

Using the indexing and slicing we add the ROI to the original images.
img1[0:min_height,0:min_width] = img_add
img2[0:min_height,0:min_width] = img_add

Visualize the added images

plt.imshow(img1)
plt.show()
Output:
Added image1 and Image2
Added Image (Image1+image2)

We have added the images. Notice the second image is added in the top-left part of the first image. Can you find the reason of this?
Can you add the second image image in bottom- right part of the first image or any other part?
For these questions look at FAQs at the end of the post.

See how the second image looks after adding. 
plt.imshow(img2)
plt.show()
Output:
Second Image after addition
Added Image

Complete Program 1 (add images of different sizes)

Now look at the complete program to add images of different size using OpenCV Python. The program describes the steps above discussed.
# import required libraries
import cv2
import matplotlib.pyplot as plt
# Read two images
img1 = cv2.imread('Desert.jpg')
img2 = cv2.imread('Penguins.jpg')


def add_images(img1, img2):
 
    height1, width1 = img1.shape[0],img1.shape[1]
    height2, width2 = img2.shape[0],img2.shape[1]
     
    min_height = min(height1, height2)
    min_width = min(width1, width2)
 
    img11 = img1[0:min_height, 0:min_width]
    img22 = img2[0:min_height, 0:min_width]
 
    img_add = cv2.add(img11, img22) 
 
    img1[0:min_height,0:min_width] = img_add
    return img1


img_add_result = add_images(img1, img2)


# cv2.imshow('added image',img_add_result)
# cv2.waitKey(0)
# cv2.destroyAllWindows()


# Convert the image from BGR to RGB
img_add_result = cv2.cvtColor(img_add_result, cv2.COLOR_BGR2RGB)
plt.imshow(img_add_result)

plt.show()
Finally we have completed all our task listed in the Table of Contents. 
Now look at some FAQS

FAQs

Q1. How to add smaller image at bottom-right part of bigger image.

Answer:

Please make changes in the step- Finally add ROI to original Images. You can make indexing and slicing accordingly. Have a look at below code snippet.
img1[height1-min_height: height1, width1-min_width:width1] = img_sub
You can add this modified line in the complete program 1 and see how the output looks like.
# import required libraries
import cv2
import matplotlib.pyplot as plt
# Read two images
img1 = cv2.imread('Desert.jpg')
img2 = cv2.imread('Penguins.jpg')


def subtract_images(img1, img2):
    
    height1, width1 = img1.shape[0],img1.shape[1]
    height2, width2 = img2.shape[0],img2.shape[1]
    
    min_height = min(height1, height2)
    min_width = min(width1, width2)
    
    img11 = img1[0:min_height, 0:min_width]
    img22 = img2[0:min_height, 0:min_width]
    
    img_sub = cv2.subtract(img11, img22)    
    
    img1[height1-min_height: height1, width1-min_width:width1] = img_sub
    
    return img1


img_sub_result = subtract_images(img1, img2)


cv2.imshow('Subtracted image',img_sub_result)
cv2.waitKey(0)
cv2.destroyAllWindows()


# Convert the image from BGR to RGB
img_add_result = cv2.cvtColor(img_sub_result, cv2.COLOR_BGR2RGB)
plt.imshow(img_add_result)
plt.show()
Output:
Smaller image is added at bottom-right of bigger image
Added image at bottom-right


Q2. How to add smaller image at bottom-left part of bigger image.

Answer:

As in the first Question, please make changes in the step- Finally add ROI to original Images. You can make indexing and slicing accordingly. Have a look at below code snippet.

img1[height1-min_height: height1, 0:min_width] = img_sub
You can add this modified line in the complete program 1 and see how the output looks like.
Smaller image is added at bottom-left of bigger image
Added image at bottom-left


Q3. How to add smaller image at left-bottom part of bigger image.

Answer:

As in the first Question, please make changes in the step- Finally add ROI to original Images. You can make indexing and slicing accordingly. Have a look at below code snippet.
img1[0:min_height, width1-min_width:width1] = img_sub
You can add this modified line in the complete program 1 and see how the output looks like.
Smaller image is added at top-right of bigger image
Added image at top-right

We have covered all the content related to adding and subtracting images of different sizes ( different height and width). If you have any doubts or want more related to this topic please comment.


Advertisements







Comments

Post a Comment