Image Forgery Identification using Convolution Neural Network
N. Hema Rajini

N. Hema Rajini, Department of Computer Science and Engineering, Alagappa Chettiar Government College of Engineering and Technology, Karaikudi (Tamil Nadu), India.
Manuscript received on 03 June 2019 | Revised Manuscript received on 28 June 2019 | Manuscript Published on 04 July 2019 | PP: 311-320 | Volume-8 Issue-1S4 June 2019 | Retrieval Number: A10550681S419/2019©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (

Abstract: In recent days, an important problem in image forensic is to determine whether a specific image is authenticated or non-authenticated. It is a crucial process where the images are considered as major evidence to alter decision in various scenarios like court laws. For performing those forensic investigation, different technical devices and algorithms has been introduced in the present world. Copy move and splicing are the commonly employed approaches for passive image forgery. This paper develops a model for detecting splicing and copy-move forgery concurrently on the similar dataset of CASIA v1.0 as well as CASIA v2.0. At the beginning, a doubtful image is considered for processing and the feature extraction process takes place using block discrete cosine transform (BDCT) and enhanced threshold approach. The presented model will decide the presence of manipulated image among the provided images. When the image is found to be manipulated, convolution neural network (CNN) is employed for classifying the image into splicing or copy-move forgery. Furthermore, Zernike Moment (ZM) polar is employed for locating the replicaportions in the image. The simulation outcome ensures the effective performance of the presented method over the existing ones.
Keywords: Image Forensic, Copy Move, CNN, Splicing, Zernike Moment.
Scope of the Article: Image Processing and Pattern Recognition