Frequency domain manipulation of multiple copy-move forgery in digital image forensics

Area of Science:

• Digital Image Processing
• Computer Vision
• Forensic Science

Background:

• Copy-move forgery involves duplicating image regions, posing detection challenges due to consistent image properties.
• Conventional and deep learning methods struggle with copy-move forgeries involving transformations like resizing.
• Robust detection of transformed copy-move forgeries remains an open research problem.

Purpose of the Study:

• To propose a novel transform domain method for generating and analyzing multiple copy-move forgeries.
• To address the limitations of existing methods in detecting forgeries with transformations.
• To evaluate the proposed method's effectiveness against state-of-the-art techniques.

Main Methods:

• Utilizes the discrete wavelet transform (DWT) to decompose images into approximate and detail coefficients.
• Simulates multiple copy-move forgeries by embedding transformed patch coefficients into original image coefficients.
• Employs the inverse DWT (IDWT) for reconstructing the forged image for analysis.

Main Results:

• The proposed method successfully generates multiple copy-move forgeries with transformations.
• Evaluation against existing techniques highlights the method's potential for improved detection.
• Varying patch sizes and transformations yield diverse outcomes, informing further research.

Conclusions:

• The transform domain approach offers a promising avenue for creating and analyzing complex image forgeries.
• Identified gaps in current detection techniques necessitate further development of robust methods.
• This work contributes to advancing the field of digital image forensics.