Key-point based copy-move forgery detection in digital images

Abstract

Nowadays all the visual forms have become digital for the communication of important and useful data. It has become imperative to check the authenticity of the digital content available in the form of images. Image forensic is an emerging area to check the realness and reliability of digital images. Amongst various image forgeries known, copy-move forgery poses a serious threat to the society and image forensic experts. “copy-move” is considered one of the most difficult problems in the area of forgery detection in image forensics. In copy-move forgery, a partof its image content is copied and pasted within the same image. For the detection of copy move image forgery, many methods have been proposed such as block based method, exhaustive search, key-point based method and hybrid method. Key-point based copy-move forgery detection performs better as compared to block based methods in the presence of affine transformation and invariant to scaling, rotation and noise. Keypoint based methods are computationally efficient and have better space complexity. In this paper, a comprehensive review of the recent key-point based methods using SIFT, SURF, ORB, BRISK is presented.

Keywords:

• Copy-move image forgery
• Key-point based methods
• SIFT
• SURF
• ORB
• BRISK
• Hybrid methods

Mathematics Subject Classification 2010:

• 68U10