HOME       POLICIES       SUBMISSION        PEOPLE        ARCHIVES         CONFERENCES        CONTACT

ABSTRACT


A copy-move tampering is induced by facsimileing and gluing content inside a similar picture, and conceivably post preparing it. As of late, the identification of copy-move fabrications has turned into the most dynamic research region in blind image forensics. A many copy-move forgery detection (CMFD) techniques such as block-based and key-point matching predicated methods have been proposed for the detection and localization of copy-move image forgery. Key-point based strategies were found better than block-based methods in term of robustness against affine transformations, computational cost and space complexity, still, it has some limitations. Researchers have proposed many integrated methods to cope up with the limitations of key-point based methods, but they failed to detect mirror reflection transformation and some improvements to define the tampered region more precisely is required. This paper aims to propose a novel method which can elevate these drawbacks by combining features of key-point based method MIFT (Mirror Invariant Feature Transform) and block-based technique Zernike moments. Additionally, proposed refining the affine transformation utilizing an iterative plan which amends the estimation of the affine transformation parameters by incrementally finding extra key-point matches. The proposed integrated method has been assessed utilizing a sizably voluminous dataset of real images. The improved results are obtained compared with MIFT and Fusion method for parameters precision, recall, and F-measure.


KEYWORDS

Image forensics, Copy-Move Image Forgery, SIFT, MIFT, Zernike Moments.
ARCHIVES
To return to the Volume/Number webpage, click here.
THE INTERNATIONAL JOURNAL OF FORENSIC COMPUTER SCIENCE - IJoFCS

Volume 13, Number 1, pages 29-36, DOI: 10.5769/J201801003 or http://dx.doi.org/10.5769/J201801003


Comprehensive study of digital forensics branches and tools


By  Lucas Oliveira Batista, Gabriel Adriano de Silva, Vanessa Souza Araújo, Vinícius Jonathan Silva Araújo, Thiago Silva Rezende, Augusto Junio Guimarães, Paulo Vitor de Campos Souza



To download this paper, click here.