A Passive Technique for Image Forgery Detection in Digital Images

Abstract

Digital images are simple to operate and edit due to accessibility of much software. It is feasible to insert or eliminate some features from an image. Cloning is one of the image tampering methods, replicating the same region of the image. The proposed method works in finding the cloned and uncloned forgeries of newly added region to the real image wherein a region from an image is replaced with another region from the different image. The existing techniques in finding alike regions suffer from their lack of ability to spot the cloned region when it has been subjected to a geometric transformation. The Proposed method clearly employ the distortion cues can identify the forgery region in distortion image. Two bottom-up cues are proposed based on distortion constraint are provide to differentiate the validation of the line in the image. Then a fake saliency map is used to maximum fake detection density, and based on the fake saliency map, an energy function is provided to get the pixel-level forgery object.