Seamless Cloning in Image Processing

Types of Seamless Cloning

Some of the common types of Seamless Cloning are

Poisson Image Editing

Poisson Image Editing is a type of seamless cloning that is used for producing smooth and visually pleasing images. It is like a magical tool that helps make the picture look better or different in a very smooth and natural way. 

For example, 

Imagine you have a photo, and you want to change the colour of the sky. Now, instead of just pasting another sky to replace the original sky, Poisson editing carefully blends the two together. It does this so that the colours and details match perfectly, and it looks like the new sky is always there.

Texture Synthesis

Texture synthesis is a smart computer technique for creating realistic-looking textures or patterns. It's like a digital artist who can make beautiful designs, like the texture of wood or the pattern on a rug, without needing to draw every bit by hand.

For example,

Imagine you liked the texture of the carpet in an image showing the room. This small piece is called a sample, This technique cleverly blends the edges of each copy of the sample so they smoothly transition into each other. As a result, the texture gradually gets bigger and bigger.

Gradient Domain Blending

Gradient Domain Blending is another type of Seamless Cloning. It is used to combine two pictures smoothly in a way to create one pleasing image. It merges the two images together like they belong together naturally. 

For example,

You have two an image of your mom and dad and one more image of your sister with your grandmother and grandfather. Now, you want to merge the two images to create a single-family picture. Instead of just sticking the two photos together, it makes sure that the gradual changes in colours look smooth and natural. It's like mixing paint colours to get the perfect shade.

Exemplar-Based Cloning

Exemplar-based cloning is like editing a picture to change or improve it but in a smart and realistic way. It includes copying and pasting sections of one picture into another picture and making sure it looks original.

For example,

Imagine you have a picture of a stadium, and you very cleverly want to place your image in the audience, but you never went to watch the match. Instead of just pasting your picture into the original picture, exemplar-based cloning looks closely at the surroundings and makes sure that the new part blends in perfectly. Thus making it look original.

Content-Aware Fill

Content-Aware Fill is another technique that helps to remove unwanted objects from an image. It’s like having a photo magician on your computer. So, Content-Aware Fill is a fantastic tool for retouching photos and images, making them look more polished.

For example,

Imagine you have taken a picture in a garden, and some random person has ruined your image by walking behind you. You can easily remove the person using this tool. It fills in the selected area with something that matches the surroundings. It's like the object was never there in the first place.

Implementation of Seamless Cloning

OpenCV provides a convenient function for performing seamless cloning. Let’s see how we can use it. Before starting with the implementation, let’s see the resources we need.

Pre-Requisite

We will require two images to perform cloning. You can download these two images if you want to perform the implementation with us.

Image 1:

Let’s try to clone this football on the above field.

Image 2: 

Steps of Implementation

Let’s see the steps of implementation of seamless cloning.

• Image Loading: Load both images using the imread function. The first dst image represents the destination image, and the second src image represents the source image.
   
• Creating a mask: A white mask is created with the same size and data type as the source image. This mask will help define which part of the source image will be cloned into the destination image.
   
• Center Calculation: The center of the image is calculated by first determining the dimensions of the image using im.shape, and the center variable is used to set the center point of the image.
   
• Seamless Cloning: Seamless cloning is performed using the cv2.seamlessClone function. The blending mode is specified using the cv2.NORMAL_CLONE and cv2.MIXED_CLONE.
  
  
  • In normal cloning, the source image replaces the corresponding region in the destination image. It is suitable for scenarios where you want to completely replace a region in the destination image.
     
  • In mixed cloning, it blends the two images in a way that preserves the texture of the source image while adapting to the lighting and colours of the destination image. It is more flexible than normal cloning.
     

Code

• Python

Python

# import libraries
import cv2
import numpy as np

# Read images
dst = cv2.imread("image1.jpg")
src = cv2.imread("image2.jpg")

# Create an all white mask
mask = 255 * np.ones(src.shape, src.dtype)

# The location of the center of the src in the dst
height, width, channels = dst.shape
center = (width // 2, height // 2)

# Seamlessly clone src into dst and put the results in output
normal_clone = cv2.seamlessClone(src, dst, mask, center, cv2.NORMAL_CLONE)
mixed_clone = cv2.seamlessClone(src, dst, mask, center, cv2.MIXED_CLONE)

# Write results in output image
cv2.imwrite("output_normal.jpg", normal_clone)
cv2.imwrite("output_mixed.jpg", mixed_clone)

You can also try this code with Online Python Compiler

Run Code

Output

Normal Cloning

Mixed Cloning

Applications of Seamless Cloning

Seamless Cloning is used in many fields, The following are some of the applications of seamless cloning.

Photo Editing

Seamless cloning is widely used in photo editing software like Adobe Photoshop to remove unwanted objects from the images. 

Image Restoration

In historical or damaged photos, seamless cloning is widely used to restore missing or damaged parts, like torn edges, while preserving the original texture and appearance.

Medical Imaging

Seamless cloning can be used in the medical field to remove artefacts or labels from radiological images while preserving the underlying anatomical structures.

Film Industry

The Film industry mainly uses seamless cloning to add or remove elements from the scenes. It is used for creating special effects, like fighting scenes, jumping from buildings, etc.

Digital Art

Seamless cloning allows for the creation of visually stunning and imaginative artworks by blending multiple images together.

Frequently Asked Questions

What is the main purpose of Seamless Cloning?

The main aim of seamless cloning is to enhance the images for various different applications. It contains removing unwanted objects, improving lights and exposure, creating composite images, etc.

What are some of the challenges in Seamless Cloning?

The challenges of seamless cloning include handling variations in lighting and texture, maintaining the quality of images, ensuring scalability when working with large images, and many more.

What software tools are commonly used for seamless cloning?

The seamless cloning can be performed using both editing software and libraries and frameworks, like Adobe Photoshop, OpenCV in Python, GNU Image Manipulation, etc.

Conclusion

In this article, we discussed, in brief, Seamless Cloning in Image Processing and its types and examples. We have also performed an implementation of Seamless cloning in OpenCV. At last, we discussed some applications of it.

To learn more about Image Processing, you can read these articles.

• Image Processing with Python
• Image Sampling and Quantization
• Smoothing Images
  
   

Refer to our Guided Path to upskill yourself in DSA, Competitive Programming, JavaScript, System Design, and many more! If you want to test your competency in coding, you may check out the mock test series and participate in the contests hosted on Coding Ninjas Studio!

But suppose you have just started your learning process and are looking for questions from tech giants like Amazon, Microsoft, Uber, etc. For placement preparations, you must look at the problems, interview experiences, and interview bundles.

We wish you Good Luck! 

Happy Learning!