A Crypto-attacking Scheme for a (t, n)-threshold Image Encryption Method

Recently, (t, n)-threshold secret sharing technology has found applications in reversible data hiding in encrypted images (RDHEI) for image encryption. It offers advantages such as independence from key management systems and resilience against n − t points of failure. In [15], Hua et al. introduced the matrix secret sharing (MSS) method, which utilizes matrix multiplication for secret sharing. Building upon this foundation, MSS has been further applied to image encryption for RDHEI methods. In our study, we conduct a comprehensive analysis of the properties of the MSS-based image encryption scheme. Subsequently, we propose an innovative cryptanalysis approach based on vector quantization (VQ) attacks. Specifically, we introduce the concept of an element variance sequence to capture the changing patterns of the ciphertext image block-wise. We then employ VQ technology to estimate the plaintext content based on the captured changing patterns. Our experimental results provide evidence for the effectiveness of our proposed cryptanalysis approach. Based on this study, we conclude that the MSS scheme raises security concerns.