Image forgery detection and localization using modified JPEG ghost

Abstract

Authenticity of an image taken digitally suffers severe threats as a result of increase in various powerful digital image editing tools. These tools modifies the image contents without leaving footprint of such modifications. We come up with a technique that analyzes digital image forgery detection in JPEG images which goes through multiple compression. Nearly all digital devices uses JPEG as a standard storage format to maintain the storage space. JPEG is a lossy compression standard. By using any image processing tools, when assailant changes any part of a JPEG image and save it, the alter part of the image has different compression artifacts. JPEG ghost algorithm is used to detect disparity in JPEG blocks that rise from improper alignments of JPEG blocks respect to original structure and detect local footprint of JPEG compression. In our work, our proposed technique will modify JPEG ghost detection to detect and localize digital image forgery.