Replication regions detection in images based on frequency analysis of blocks

Nowadays, image processing software has been improved progressively. Image modifications with unnoticed tricks are available via advanced software. There are many image tricky techniques and Copy-Move Forgery is the most extensive. It is called copy-move forgery when duplicated image region is pasted to another location in same image. In this study, image is divided into sub-blocks to detect tricky and forgery. Contribution to literature is that feature vectors of the blocks are obtained using combination of Fourier transform of rows and columns. Similarities of the duplicated regions are searched through Fourier transform of the blocks. It seems that reduced feature vectors represent effectively the image blocks even though size of feature vectors is decreased. Accuracy of the proposed algorithm is investigated under different distortions. Results given in tables and figures prove that the suggested method can detect successfully duplicated regions under JPEG compression corruption and Gaussian Blurring disruption.