Image Forgery Detection Using Dct And Quantization Matrix Techniques
Shiji Abraham1, Anisha P Rodrigues2, Roshan Fernandes3

1Shiji Abraham, Department of Computer Science and Engineering, NMAM Institute of Technology, Nitte, India.
2Anisha P Rodrigues, Department of Computer Science and Engineering, NMAM Institute of Technology, Nitte, India.
3Roshan Fernandes, Department of Computer Science and Engineering, NMAM Institute of Technology, Nitte, India.
Manuscript received on July 20, 2019. | Revised Manuscript received on August 10, 2019. | Manuscript published on August 30, 2019. | PP: 4575-4581 | Volume-8 Issue-6, August 2019. | Retrieval Number: F8883088619/2019©BEIESP | DOI: 10.35940/ijeat.F8883.088619
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: In today’s era the image has become useful for communication purpose. But due to the development of software and various techniques it is possible to change images in adding or removing essential feature from it without leaving a clue of real image. It is not easy for the common people to identify whether the image original or tampered. In order to avoid this problem, forgery detection came into existence. Detection of forgery refers to task of image processing to identify that the images are unique or tampered. Several techniques have been used in order to detect the forgeries from the forged image, but this issue has not yet solved. In order to solve these issues we have used Discrete Cosine Transformation (DCT) and quantization matrix techniques for identifying forged areas of image, where the quality of image is not reduced. The Discrete Cosine Transformation (DCT) is used in order for characterizing the overlapping blocks and quantization matrix is used to compress DCT values and gives both highly compressed and best decompressed image quality. Here we use block matching algorithm. This algorithm one of the most frequently used for detecting image which is duplicate. This proposed work also supports for different kinds of images such as JPEG, JPG or PNG of any size it can be either mxn or nxn.
Keywords: Forgery detection, image processing, tampered image, DCT based algorithm.