Quick response (QR) codes have become increasingly popular as a medium for quickly and easily accessing information through mobile devices. However, the open-source nature of QR code encoding poses a risk of information leakage and potential attacks, especially with the growing use of QR codes in financial services and authentication applications. To mitigate the risk of information leakage due to open-source QR code encoding, this paper proposes a two-level QR code scheme based on a region matrix image secret sharing algorithm. In this scheme, the first-level public information can be directly obtained by scanning with any standard QR code scanner, while the two-level secret information can only be accessed by overlaying the shared images. To enhance the robustness of joint secret information recovery using shared images, this article designs a progressive image secret sharing algorithm based on region matrices. This algorithm meticulously processes high-priority share regions and generates multiple substitute shares. In the event of attacks on key shares, substitute shares can be employed to recover the secret information. For an increased payload capacity of each QR code, an adaptive pixel depth adjustment algorithm is devised. This algorithm ensures that the recovery of two-level secret information maintains high clarity, while not affecting the scanning functionality of each shared QR code. Experimental results validate the feasibility of this scheme, which simplifies the construction matrix, reduces matrix redundancy, and exhibits priority partitioning and higher robustness. Furthermore, QR codes embedding secret shares can safeguard the two-level information, and the recovered images exhibit exceptional clarity.
Keywords: QR code; robustness; two-level QR code; visual secret sharing.