Arithmetic Operations on Images using OpenCV

Arithmetic operations such as addition, subtraction and bitwise operations (AND or, NOT, XOR) are fundamental techniques in image processing with OpenCV. These operations allow for the enhancement, analysis and transformation of image characteristics, making them essential for tasks like image clarification, thresholding, dilation and more.

Let's see the step by step implementation of Arithmetic operations,

Step 1: Install and Import Required Libraries

Before performing arithmetic operations, install and import the required libraries.

!pip install opencv - python matplotlib

import cv2
import numpy as np
import matplotlib.pyplot as plt
from google.colab import files

Explanation:

• opencv: Core library for image processing and computer vision.
• matplotlib.pyplot: For displaying images inside the notebook .
• numpy: Efficient array operations .

Step 2: Load the Input Images

Next, load the images on which the arithmetic operations will be performed. Make sure both images are available in the current working directory.

Note: The samples used can be downloaded from here.

img1 = cv2.imread('input1.png')
img2 = cv2.imread('input2.png')

Explanation:

• files.upload() opens a dialog to pick files from our device.
• cv2.imread() reads an image from disk and loads it as a NumPy array (in BGR color ordering by default).

Step 3: Display the Input Images

Before applying arithmetic operations, it is useful to display both images side by side for comparison. If the images have different dimensions, resize one image so that both have the same size.

if img1.shape != img2.shape:
    img2 = cv2.resize(img2, (img1.shape[1], img1.shape[0]))

line_thickness = 5
height = img1.shape[0]
line = np.full((height, line_thickness, 3), (0, 0, 255), dtype=np.uint8)

side_by_side = np.hstack((img1, line, img2))

side_by_side_rgb = cv2.cvtColor(side_by_side, cv2.COLOR_BGR2RGB)

plt.figure(figsize=(12, 6))
plt.imshow(side_by_side_rgb)
plt.title('input1 input2')
plt.axis('off')
plt.show()

Output

Explanation:

• img1.shape != img2.shape checks whether the two images have different dimensions.
• cv2.resize() resizes the second image to match the dimensions of the first image.
• np.full() creates a vertical red line that acts as a separator between the images.
• np.hstack() combines both images horizontally into a single image.
• cv2.cvtColor() converts the image from BGR to RGB format for correct display with Matplotlib.
• plt.imshow() displays the combined image, allowing you to visually compare both inputs before applying arithmetic operations.

Step 4: Apply Arithmetic Operations

1. Image Addition

Image addition combines the pixel values of two images. OpenCV automatically handles overflow by limiting pixel values to the valid range (0–255).

Simple Addition: cv2.add() adds corresponding pixel values from both images.

added = cv2.add(img1, img2)
added_rgb = cv2.cvtColor(added, cv2.COLOR_BGR2RGB)

plt.imshow(added_rgb)
plt.title('Addition (cv2.add)')
plt.axis('off')
plt.show()

Output

Explanation:

• cv2.add() adds pixel values from both images.
• Values greater than 255 are clipped to 255.
• Commonly used for increasing image brightness or combining images.

2. Weighted Addition

Weighted addition blends two images by assigning a contribution weight to each image.

weighted = cv2.addWeighted(img1, 0.7, img2, 0.3, 0)
weighted_rgb = cv2.cvtColor(weighted, cv2.COLOR_BGR2RGB)

plt.imshow(weighted_rgb)
plt.title('Weighted Addition (cv2.addWeighted)')
plt.axis('off')
plt.show()

Output

Explanation:

• 0.7 determines the contribution of img1.
• 0.3 determines the contribution of img2.
• The last parameter (0) is the gamma value used for brightness adjustment.
• Frequently used for image blending and transparency effects.

3. Image Subtraction

Image subtraction removes pixel values of one image from another.

subtracted = cv2.subtract(img1, img2)
subtracted_rgb = cv2.cvtColor(subtracted, cv2.COLOR_BGR2RGB)

plt.imshow(subtracted_rgb)
plt.title('Subtraction (cv2.subtract)')
plt.axis('off')
plt.show()

Output

Explanation:

• cv2.subtract() subtracts each pixel in img2 from the corresponding pixel in img1.
• Negative values are automatically clipped to 0.
• Commonly used for change detection and background subtraction.

4. Bitwise AND

Bitwise AND keeps only the pixels that are active in both images.

and_img = cv2.bitwise_and(img1, img2)
and_img_rgb = cv2.cvtColor(and_img, cv2.COLOR_BGR2RGB)

plt.imshow(and_img_rgb)
plt.title('Bitwise AND')
plt.axis('off')
plt.show()

Output

Explanation: Pixels are retained only when corresponding bits are set in both images, which is commonly used for masking operations.

5. Bitwise OR

Bitwise OR keeps pixels that are active in either image.

or_img = cv2.bitwise_or(img1, img2)
or_img_rgb = cv2.cvtColor(or_img, cv2.COLOR_BGR2RGB)

plt.imshow(or_img_rgb)
plt.title('Bitwise OR')
plt.axis('off')
plt.show()

Output

Explanation: A pixel is retained if it exists in either image. Useful for combining image regions.

6. Bitwise XOR

Bitwise XOR keeps pixels that differ between the two images.

xor_img = cv2.bitwise_xor(img1, img2)
xor_img_rgb = cv2.cvtColor(xor_img, cv2.COLOR_BGR2RGB)

plt.imshow(xor_img_rgb)
plt.title('Bitwise XOR')
plt.axis('off')
plt.show()

Output

Explanation: Pixels are retained only when the corresponding bits are different. Useful for highlighting differences between images.

7. Bitwise NOT

Bitwise NOT inverts all pixel values in an image.

not_img = cv2.bitwise_not(img1)
not_img_rgb = cv2.cvtColor(not_img, cv2.COLOR_BGR2RGB)

plt.imshow(not_img_rgb)
plt.title('Bitwise NOT (Image 1)')
plt.axis('off')
plt.show()

Output

Explanation: cv2.bitwise_not() inverts every pixel value and dark pixels become bright and bright pixels become dark. Often used for image inversion and mask generation.