Neighbourhood Projection Embedding Based Image Tampering Detection and Localization

Image tampering detection is a well developed field that not only analyzes the authenticity of image but restores credibility of it. Motive of image tampering is to create false notion about an image in viewers prospective. Image can be forged by modifying various features of image or by adding or eliminating part of it. Copy-move tampering can be defined as the process of inserting or deleting image region from an image such that no proof of alteration is visible. In this paper we propose to use Neighbourhood Projection Embedding (NPE) with regard to detection of copy-move tampering. Neighbourhood information preserving property of NPE can effectively detect and localize tampering in the presence of various post-processing operations like, additive Gaussian noise, JPEG compression, brightness change, colour reduction and blur images effectively.