Copy-move forgery detection algorithm using frequency transforms, surf and mser

The number of digital images shared in the internet is growing faster, increasing the risk that these images be modified causing moral or economic damage. Thus, it is necessary to develop efficient image authentication algorithms that be able to detect malicious modification on the digital images. To solve this problem a tampering detection algorithm is proposed to detect if a digital image was tampered, using the Copy-Move technique. In the proposed scheme the input image is segmented into N×M blocks whose similarity is analyzed using frequency domain techniques, together with SURF or MSER detectors. The evaluation results show the desirable features of proposed approach.