View article

Copy-move forgery is a specific type of image tampering where a part of the image is copied and pasted on another part generally to conceal unwanted portions of the image. Hence, the goal in detection of copy-move forgeries is to detect image areas that are same or extremely similar. In this paper, we review several methods proposed to achieve this goal. These methods in general use block-matching procedures, which first divide the image into overlapping blocks and extract features from each block, assuming similar blocks will yield similar features. Later, a matching step takes place where the aim is to find the duplicated blocks based on their feature vectors. A forgery detection decision is made only if similar features are detected within the same distance of features associated to connected blocks. We examine several different blockbased features proposed for this purpose in relation to their time complexity and robustness to common processing scaling up/down, compression, and rotation. We will also include comparative results for better evaluation.