Abstract :

Abstract- A Common Method Of Image Modification Is Known As Copy-move Forgeries. This Method Involves Copying And Pasting A Portion Of An Image Onto Another Location, Typically With The Intention Of Concealing Or Changing Information. Within The Realm Of Digital Media Forensics, It Is Frequently Encountered, And Its Applications Include The Detection Of Images That Have Been Altered, The Verification Of Authenticity, And The Maintenance Of Integrity In Legal And Journalistic Contexts. At The Moment, The Detection Of Copymove Forgery Is Primarily Dependent On Manual Analysis Performed By Forensic Specialists. The Procedure Is Visually Evaluating Photographs, Searching For Abnormalities In The Patterns, Lighting, And Textures Of The Images. Manual Analysis, Despite Its Dependability, Is Time-consuming And Resource-intensive, Which Limits Its Scalability And Efficiency In The Management Of Huge DatasetsIn The Proposed Copymove Forgery Detection System, VGG16 Is Utilized As A Feature Extractor To Identify Patterns Indicative Of Tampered Regions, Leveraging Its Hierarchical Convolutional Layers To Capture Both Global And Local Inconsistencies In Images. The Extracted Deep Features Are Then Processed Using DBSCAN Clustering, Which Segments The Image By Grouping Similar Feature Points And Isolating Potential Forged Areas Based On Density. This Combination Enhances Forgery Detection By Effectively Identifying Duplicated Regions While Filtering Out Noise. The Detected Regions Are Then Refined Using Morphological Operations, And The Results Are Visualized By Overlaying The Forgery Map On The Original Image. This Approach Ensures Robust And High-precision Detection Of Manipulated Content, Making It Highly Effective For Image Authentication Tasks. Keywords: Copy-move Forgery, Digital Image Forensics, Transfer Learning, VGG16, Feature Extraction, DBSCAN Clustering, Image Tampering Detection, Morphological Operations, Forgery Localization, Image Authentication.

Published:

15-5-2025

Issue:

Vol. 25 No. 5 (2025)

Page Nos:

455-459

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite