Copy-move Forgery Detection using SIFT and DWT detection Techniques

Abstract

Copy-Move forgery practices are defamatory tampering processes, affecting digital images. Such practices have become much more common nowadays because of the advent of significantly effective image processing tools being readily available. In copy-move forgery technique, an area from the original image is copied and pasted over another image area. This technique provides a way to conceal the important details and information in a digital image without leaving much traces behind to observe the tampering or manipulation of the image. Such practices question the authenticity of an image to forensic. In this paper, the techniques are proposed for the detection of cloning or copy-move forgery that involve both block based extraction using Discrete Wavelet Transform (DWT) and feature point extraction based detection technique i.e., Scale Invariant Feature Transform (SIFT) to extract the tampered regions more precisely. The proposed algorithm mainly involves tentacle matching of features of same features, which can be extracted from each of the blocks by calculating the dot products between their respective unit vectors. The proposed algorithm is able to achieve an efficiency of 98.12%, precision factor of 97% of and recall factor of 100%.