Detection of Near-Duplicated Image Regions

Modern, easy to use image processing software enables forgeries that are undetectable by the naked eye. In this work we propose a method to automatically detect and localize near-duplicated regions in digital images. The presence of nearduplicated regions in an image may signify a common type of forgery called copy—move forgery. The method is based on blur moment invariants, which allows successful detection of copy—move forgery, even when blur degradation, additional noise, or arbitrary contrast changes are present in the duplicated regions. These modifications are commonly used techniques to conceal traces of copy—move forgery. Our method works equally well for lossy format such as JPEG.