How do I match clusters within an image to find copy-move forgery

I am trying to implement this paper on copy-move forgery in matlab. I have created clusters using selected keypoints. And now I need to identify the matching clusters and mark them with each other . But I could not find any understandable method to match the clusters. It would be helpful if someone could advice me on this. As an example, I extracted and marked clusters in the below image

Answer 1

[Image Analyst]

The Computer Vision System Toolbox has SURF, which is like SIFT. You'll just have to figure out how to implement the paper with the function. See https://www.mathworks.com/help/vision/ug/local-feature-detection-and-extraction.html?searchHighlight=sift&s_tid=doc_srchtitle

[Kevin Kevin]

@Image Analyst: Thank you for the quick response.

I have gone through the given link. And I believe, that is all about matching the keypoints. Please note, my implementation is for academic purpose.

My concern is, matching the keypoints alone, won't cover a region(or a specific area). So what I have done is(according to the paper),

1)Extracted the keypoints from the image(using SIFT)

2)Separated the matching keypoints(in order to avoid, creation of unnecessary clusters, thus to reduce computation)

3)Extracted clusters from the matching keypoints(please refer the image in question)

Finally, I need to connect those matching clusters.

In this case(as I understand it. Pardon, if I'm wrong), clusters are generated to highlight the 'regions' that are forged.

[Image Analyst]

OK, good luck. If it ends up being too much work, maybe you can see if the authors will sell you their software that does it already, and save you a lot of time.

[Kevin Kevin]

@Image Analyst: Thank you for the suggestion. :)