A DWT-DCT Based Passive Forensics Method for Copy-Move Attacks

A passive forensics method for detecting copy-move attacks was proposed. In the proposed method, tested image was segmented into non-overlapping blocks, DCT and DWT are applied to each image block to extract features, and then compare the statistics of each image block to detection replication regions. Our method was robust to compression up to JPEG quality level 20 and against additive noise. The time complexity of our approach for sorting is O(k 1g k).