Image Forgery Detection Using an Adaptive and Improved Methodology

Authors

  • Mohammad al-Kurdi Syrian Virtual University
  • Raneem Tasseh Syrian Virtual University

Keywords:

Splicing Forgery; Cloning Forgery; DCT; SURF; CFA.

Abstract

The study presented through this research deals with the subject of digital images, it is specifically directed to detect forgery that can be applied to digital images. The proposed system provides an adaptive and dynamic application, which can detect the two most common and used types of forgery: Splicing forgery, and Cloning (copy-move) forgery. This system can detect these two types of forgery, in different types and sizes of images. Unlike many previous studies that were intended for a specific type of forgery, or for an image with specific characteristics and conditions. The application dynamically adapts to the given image and selects the optimal algorithm that fits image, so to arrive to the best result in detecting the forgery as perceived by the image data and specifications.
      The proposed system also provides improvement concerning the number of false alarms (False Positive), that were issued by the basic systems that detect Cloning forgery. As the two basic systems presented in previous studies were suffering from a large number of false alarms, that showed the existence of fraud while the original image was not forged. Therefore, one of the aims of this study was to search for the causes of these false alarms in each method separately, and to treat these causes so to improve the performance of the original algorithms

Author Biographies

Mohammad al-Kurdi, Syrian Virtual University

Professor, Syrian Virtual University

Raneem Tasseh, Syrian Virtual University

Postgraduate Student (Master), Department of Web Science

Published

2021-03-10

How to Cite

1.
الكردي م, طاسية ر. Image Forgery Detection Using an Adaptive and Improved Methodology. Tuj-eng [Internet]. 2021Mar.10 [cited 2024Dec.23];43(1). Available from: https://journal.tishreen.edu.sy/index.php/engscnc/article/view/10444