A new block-based copy-move forgery detection method in digital images

Nowadays, digital image forgery detection is one of the most challenging tasks. In this paper, a new block-based method is proposed to detect tampered region in copy-move forgery case. In the method, feature vectors are extracted using DCT transform from non-overlapping blocks of the image. Then, these vectors are lexicographically sorted. Copied blocks are selected from identical vectors based on some criteria. A novelty of this paper is considering a dispersion threshold to eliminate large flat parts of the image (such as Sky) from forgery detection. Simulation results show that, the method can effectively detect forgery in the image with very small False-Positive in comparison with the classic methods in terms of FPR and DAR.