Hard EXIF: Protecting Image Authorship Through Metadata, Hardware, and Content.

Area of Science:

• Computer Science
• Digital Forensics
• Information Security

Background:

• Digital image proliferation necessitates robust authorship protection.
• Existing methods like EXIF data, watermarking, and blockchain face limitations in security, usability, and cost.
• Challenges include metadata tampering, watermarking vulnerabilities, and blockchain complexity.

Purpose of the Study:

• To introduce a novel framework, Hard EXIF, for improved digital image authorship protection.
• To address the limitations of traditional methods by balancing legal validity, security, cost, and usability.
• To integrate physically unclonable functions (PUFs) with image metadata for enhanced security.

Main Methods:

• Developed the Hard EXIF framework combining metadata with physically unclonable functions (PUFs).
• Utilized unique device fingerprints generated by PUFs and embedded them using watermarking techniques.
• Leveraged hash functions and PUFs to enhance EXIF security and minimize implementation costs.

Main Results:

• Achieved an average Peak Signal-to-Noise Ratio (PSNR) of 42.89 dB.
• Maintained high similarity (99.46%) between original and watermarked images.
• Demonstrated a low extraction error rate of 0.0017.

Conclusions:

• The Hard EXIF framework effectively balances legal validity, security, cost, and usability for digital image authorship protection.
• The integration of PUFs and metadata offers a promising, high-performance solution.
• The proposed method shows significant potential for widespread application in securing digital image ownership.