Reversible Data Hiding in Encrypted Images (RDH-EI) faces challenges in embedding capacity and security. This paper proposes an enhanced RDH-EI framework that integrates secret sharing and convolutional neural networks (CNN). The process begins by encrypting the image, followed by secret sharing, which splits the encrypted image into multiple spatially correlated shares. Data embedding is performed using a CNN, improving capacity while reducing distortion. The method ensures that the original image can be recovered without loss, even if some shares are missing or corrupted, provided enough uncorrupted shares are received. This approach is particularly useful in fields like medical imaging and secure cloud storage, where both privacy and data integrity are crucial. Experimental results show that the proposed method outperforms existing RDH-EI techniques in terms of security, data capacity, and reversibility, offering a robust solution for secure communication and storage.
Raju et al. (Thu,) studied this question.