A fast and high accurate image copy-move forgery detection approach

Copy-move is one of the most common image forgeries, wherein one or more region are copied and pasted within the same image. The motivations of such forgery include hiding an element in the image or emphasizing a particular object. Copy-move image forgery is more challenging to detect than other types, such as splicing and retouching. Keypoint based copy-move forgery detection extracts image keypoints and uses local visual features to identify duplicated regions, which exhibits remarkable performance with respect to memory requirement and robustness against various attacks. However, these approaches fail to handle the cases when copy-move forgeries only involve small or smooth regions, where the number of keypoints is very limited. Also, they generally have higher time costs owing to complex feature descriptor and more error matching points. To tackle these challenges, we propose a fast and effective copy-move forgery detection method through adaptive keypoint extraction and processing, introducing fast robust invariant feature, and filtering out the wrong pairs. Firstly, the uniform distribution keypoints are extracted adaptively from the forged image by employing the fast approximated LoG filter and performing the uniformity processing. Then, the image keypoints are described using fast robust invariant feature and matched through the Rg2NN algorithm. Finally, the falsely matched pairs are removed by employing the segmentation based candidate clustering, and the duplicated regions are localized using optimized mean-residual normalized production correlation. We conduct extensive experiments to evaluate the performance of the proposed scheme, in which encouraging results validate the effectiveness of the proposed technique, in comparison with the state-of-the-art approaches recently proposed in the literature.