Fast Rotation Invariant Detection of Region Duplication Attacks Even on Uniform Background Containing Digital Images

In copy move forgery, some area of image is copied and pasted elsewhere in the same image. This may contain two motives, either to increase number of objects or to hide objects. In this paper, we propose a fast copy—rotate-paste image forgery detection method using rotation invariant property of Zernike moments (ZM). We have focused on those images where number of objects as a result of forgery, gets increased. We performed connected component analysis before block matching to reduce time complexity and for more reliable results. The fraudster attempts to make its appearance more like original. Copied portion can be scaled, rotated or colored before pasting it. To overcome these problems ZM is used. Experimental results show that the proposed image forgery detection method performs better for rotated and scaled blocks.