A new anti-forensic scheme--hiding the single JPEG compression trace for digital image

Area of Science:

• Digital Forensics
• Image Processing
• Information Security

Background:

• Digital image forensics aims to detect forgeries by analyzing processing history and origin.
• JPEG compression leaves statistical footprints in the Discrete Cosine Transform (DCT) domain, utilized by forensic algorithms.
• Anti-forensic techniques can remove these footprints, potentially invalidating existing forensic tools.

Purpose of the Study:

• To propose a novel anti-forensic algorithm capable of concealing quantization artifacts in single JPEG compressed images.
• To evaluate the effectiveness of the proposed scheme in hiding forensic traces.
• To assess the impact of the anti-forensic method on image quality.

Main Methods:

• A chaos-based dither signal is applied to the DCT coefficients of JPEG compressed images.
• The proposed algorithm aims to remove or conceal the artifacts resulting from JPEG quantization.
• Experiments were conducted to evaluate the scheme's effectiveness and image quality loss.

Main Results:

• The proposed anti-forensic scheme effectively conceals quantization artifacts from JPEG compression.
• Experimental evaluations demonstrate the scheme's capability to mislead JPEG forensic tools.
• The method's impact on image quality was assessed and found to be acceptable.

Conclusions:

• The developed anti-forensic algorithm successfully hides digital fingerprints left by JPEG compression.
• This technique challenges the reliability of current JPEG forensic tools by obscuring evidence.
• Further research may be needed to develop robust forensic methods against such anti-forensic strategies.