International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 12 Issue: 03 | Mar 2025
p-ISSN: 2395-0072
www.irjet.net
Image Forgery Detection Techniques with a Focus on Copy-Move Forgery Problem Definition and MATLAB-Based Solutions Miss. Apeksha P. Ingle1, Dr. C.N. Deshmukh2, Dr. D.T. Ingole3 1
(Mater of Engineering , Electronics & Telecommunication Engineering) Student, Prof .Ram Meghe Institute of Techonology & Research Badnera, 2 Head of Department, Prof. Ram Meghe Institute of Techonology & Research Badnera, 3 Principal, Takshashila Institute of Eengineering and Technology Darapur, Sant Gadge Baba Amravati University, Amravati, Maharashtra ---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Digital image manipulation, specifically copy-
move forgery, poses a significant challenge given the easy access to editing software. This research examines blockoriented (including DCT, PCA, DWT) and key point-oriented (such as SIFT, SURF) detection techniques, evaluates their shortcomings, and introduces a combined methodology integrating both approaches. MATLAB-based implementation is suggested to address computational complexity, localization accuracy, and robustness to geometric transformations. A MATLAB-based solution is recommended to tackle processing efficiency, detection precision, and resilience against geometric modifications.
Key Words: : Copy-move forgery, Digital copyright, Feature Extraction, Detection.
1.INTRODUCTION
Fig. 1: Types of Image Forgery Detection
Context: The proliferation of social platforms and image manipulation tools has made forgery identification crucial for legal, media, and investigative purposes.
1.1.1
Active forgery detection methodology requires preembedded or pre-extracted data. Digital Watermarking and digital signature techniques are commonly employed methods in the active detection approach.
Challenge: Copy-move forgery, involving the duplication of content within an image, presents unique difficulties due to its subtle nature and subsequent modifications (like scaling, rotation, noise insertion).
1.1.2
Objective: Create an effective, comprehensive detection system using MATLAB to address existing methodological limitations.
Authentication verification establishes image legitimacy. Various approaches have been developed to validate authentic images. These methods can be broadly classified into two main groups:
Forgery dependent methodologies are designed to identify specific types of manipulations, such as splicing, which rely on the nature of alterations performed on the image. Forgery independent approaches detect manipulations that are not tied to specific fraud types but instead focus on traces left by processes like sharpening, blurring, and irregularities in lighting and shadow patterns.
Active techniques Passive techniques
This classification is based on whether the source image is available or not. Each approach can be further subdivided. The classification hierarchy is illustrated in figure 1.
© 2025, IRJET
|
Impact Factor value: 8.315
Passive Forgery detection Techniques
Passive techniques, also known as blind methods, utilize only the image itself for authentication and integrity verification. This approach operates on the premise that even when visual tampering signs are absent, the manipulation may disrupt underlying statistical properties through noise inconsistencies, image blur effects, enhancement artefacts, copy-move forgery, and image inpainting operations.
1.1 IMAGE FORGERY DETECTION TYPES
Active Forgery identification Techniques
|
ISO 9001:2008 Certified Journal
|
Page 251