Copy-Move Image Forgery Detection using Adhoc Algorithm

Abstract

It is easy make fake images by making use of editing software. It has become an effortless job to put together or detach some attributes from an image. Validation of digital images is very essential. Identifying the fake image is the crucial topic. In order to identify tampered image active and passive detection methods are used. An image can be tampered by using image splicing, copy-move, and retouching. In particular, copy-move attack is considered in this paper. It is essential to find out whether the image is tampered or not. An effective method for detecting forged image is proposed which uses adhoc algorithm. In this algorithm there is no need of original image as it compares the similar pixels in the given image. Clusters which are larger than block size are pulled out. Similar clusters are extracted using some similarity function. The method successfully detects the forged parts in the image and saves the forged image in JPEG format. The performance is measured on various images.