Week 1 — Copy Move Forgery Detection
Introduction
Hello world, this will become our weekly blog for semester project. We are a team of undergraduate students from Hacettepe University, Cemre Şenokur, Eflatun Doğa Aydın, Mustafa Küçük.
So… What is copy move forgery?
In the world of digitalization, what you see is most likely not real. When a photograph is not fitting to the artist had in mind, digital enhancing is the best option to make it more presentable.
One of the ways of this enhancing methods is copy move forgery, to summarize briefly copy move forgery is used to fill empty spaces or hide some unwanted parts from the original image by selecting a part of the image and pasting it to the desired place.
The detection of copy move forgery can be done by any normal algorithm by just comparing image pixel by pixel base and detecting recurring patterns of pixels. But detecting it this way is too slow and it would be hard to catch if the forged part was rotated, and this is an unreliable method for lossy compressed image file types.
Our goal in this project is finding out copy move forgeries in images by the help of our predecessors’ studies.
Dataset
We are going to use a pre-existing dataset which can be found here.