Secured Secret Sharing of QR Codes Based on Nonnegative Matrix Factorization and Regularized Super Resolution...

Area of Science:

• Computer Science
• Information Security
• Data Science

Background:

• Quick Response (QR) codes are widely used for information exchange and in healthcare IoT applications.
• The open-source nature of QR code generators and scanners makes them vulnerable to duplication and tampering.
• Existing security schemes for QR codes have limitations.

Purpose of the Study:

• To present a novel (n,n) secret-sharing scheme for secure QR code transfer.
• To enhance the security and integrity of QR code data.
• To offer an alternative to polynomial and visual cryptography-based schemes.

Main Methods:

• A novel (n,n) secret-sharing scheme based on Nonnegative Matrix Factorization (NMF).
• QR code shares are reconstructed using a regularized Super Resolution Convolutional Neural Network (SRCNN).
• Exploitation of NMF for part-based data representation and SRCNN for structural element capture.

Main Results:

• The proposed method provides a potential solution for secured exchange of QR codes with varying error correction levels.
• Experimental results demonstrate significant computational and combinatorial hurdles for adversaries attempting to compromise the system.
• Security analysis shows an adversary needs 2^58 additional share combinations and 3 × 2^88 additional computations compared to a representative approach.

Conclusions:

• The proposed NMF-based secret-sharing scheme with SRCNN reconstruction offers enhanced security for QR code data transfer.
• The method effectively addresses vulnerabilities associated with QR code duplication and tampering.
• This approach represents a significant advancement in securing information exchange via QR codes in various applications, including IoT.