Copy-Move Image Forgery Detection Using Discrete Cosine Transforms

Abstract

Digital image forgeries have led us to a situation where no digital image obtained using computing devices like PCs, smartphones and laptops are trusted to be authentic. As there are plenty of tools available free and open source, digital image forgery is no more a sophisticated job. Here we have come up with a simple but an effective technique for copy-move forgery detection in digital images as this type of forgeries is hard to be identified visually. The proposed technique uses block-based forgery detection technique. The features are extracted for each overlapping block, and then, it is compared with the features of other blocks to identify if there is forgery existing in the image. Few different images were tried, and difference in the accuracies has been observed. Accuracy varies depending on the copied image.