Keypoint based comprehensive copy‐move forgery detection

• DOI: 10.1049/ipr2.12105
• Publication Date: 2020-12-28T06:58:42Z
• Authors: Anjali Diwan; Rajat Sharma; Anil K. Roy; Suman K. Mitra
• Keywords: Image and video coding; Computer vision and image processing techniques; QA76.75-76.765; Optical, image and video signal processing; Photography; 0202 electrical engineering, electronic engineering, information engineering; Image recognition; Computer software; 02 engineering and technology; TR1-1050

Abstract Verifying the authenticity of a digital image has been challenging problem. The simplest tampering tricks is copy‐move forgery. In forgery copied portion pasted on another part same image. Geometrical transformations are used portions before pasting it for tampered to look realistic and visually convincing. To make more complex, other processing approaches may also be applied in forged region hiding traces These processings scale, rotation, JPEG compression, AWGN. this paper, an approach based features CenSurE keypoint detector FREAK descriptor proposed. This combination novelty itself as never purpose best authors' literature studies. detectors fast give stable accurate output even case rotated images, which we club with binary FREAK. Hierarchical clustering Neighbourhood search such way that can locate detect multiple forgeries. authors hopeful proposed real‐time authentication detection.