Image Arithmetic & Bitwise Operations

One of the most powerful features in OpenCV is its ability to manipulate image pixel data directly using arithmetic operations and bitwise logic. These operations form the foundation of image blending, masking, object highlighting, and more.

➕ Arithmetic Operations

OpenCV provides functions to perform element-wise addition, subtraction, division, and multiplication on images or pixel values.

cv2.add() — Safe Saturated Addition

cv2.add(src1, src2[, dst[, mask[, dtype]]]) → dst

Parameters:

• src1: First input array (image)
• src2: Second input array (image or scalar)
• dst (optional): Output image of the same size and type
• mask (optional): Operation mask
• dtype (optional): Desired output type

Unlike + operator in NumPy, cv2.add() saturates at 255 (no wraparound).

Example: Image Addition

import cv2
img1 = cv2.imread('img1.jpg')
img2 = cv2.imread('img2.jpg')
added = cv2.add(img1, img2)
cv2.imwrite('added.jpg', added)

📷 Output: Placeholder for added.jpg

cv2.subtract() — Safe Subtraction

cv2.subtract(src1, src2[, dst[, mask[, dtype]]])

• Prevents negative pixel values (saturates at 0).

subtracted = cv2.subtract(img1, img2)
cv2.imwrite('subtracted.jpg', subtracted)

📷 Placeholder for subtracted.jpg

Blending Images with Weights — cv2.addWeighted()

cv2.addWeighted(src1, alpha, src2, beta, gamma[, dst[, dtype]]) → dst

Parameters:

• src1, src2: Input images
• alpha, beta: Weight of the input images
• gamma: Scalar added to each sum

Use case: Create transparency or dissolve effect between two images.

blended = cv2.addWeighted(img1, 0.6, img2, 0.4, 0)
cv2.imwrite('blended.jpg', blended)

📷 Placeholder for blended.jpg

🔣 Bitwise Operations

OpenCV also allows for pixel-level binary logic operations:

• cv2.bitwise_and()
• cv2.bitwise_or()
• cv2.bitwise_xor()
• cv2.bitwise_not()

These are commonly used in masking and region selection.

cv2.bitwise_and()

Returns pixel-wise AND between two images (or an image and a mask).

masked = cv2.bitwise_and(img1, img2)

cv2.bitwise_or()

Returns pixel-wise OR of the input images.

combined = cv2.bitwise_or(img1, img2)

cv2.bitwise_xor()

Returns pixels where only one image has a high value.

xor_img = cv2.bitwise_xor(img1, img2)

cv2.bitwise_not()

Inverts all bits (i.e., 255 → 0, 0 → 255).

inverted = cv2.bitwise_not(img1)

🧪 Example: Masking an Object

Step-by-step Masking:

1. Load image and logo.
2. Convert logo to grayscale.
3. Threshold to create binary mask.
4. Mask the region on the base image.
5. Add the logo using bitwise OR.

import cv2
import numpy as np
 
img = cv2.imread('background.jpg')
logo = cv2.imread('logo.png')
 
rows, cols, _ = logo.shape
roi = img[0:rows, 0:cols]
 
logo_gray = cv2.cvtColor(logo, cv2.COLOR_BGR2GRAY)
_, mask = cv2.threshold(logo_gray, 10, 255, cv2.THRESH_BINARY)
mask_inv = cv2.bitwise_not(mask)
 
img_bg = cv2.bitwise_and(roi, roi, mask=mask_inv)
logo_fg = cv2.bitwise_and(logo, logo, mask=mask)
 
dst = cv2.add(img_bg, logo_fg)
img[0:rows, 0:cols] = dst
cv2.imwrite('masked_logo.jpg', img)

📷 Placeholder: Output of masked_logo.jpg

📝 Practice Questions

❓ Question 1

Blend two images using weights 0.7 and 0.3 respectively.

blended = cv2.addWeighted(img1, 0.7, img2, 0.3, 0)

❓ Question 2

Apply bitwise AND between two shaped images and save result.

result = cv2.bitwise_and(img1, img2)
cv2.imwrite('and_result.jpg', result)

📷 Placeholder for and_result.jpg

✅ Summary

• Explained cv2.add, subtract, and addWeighted
• Introduced logical bitwise operations and masking
• Detailed pixel blending for overlays
• Full masking example with binary operations