A novel reversible data hiding scheme using SMVQ prediction index and multi-layer embedding

In this paper, a novel reversible data hiding scheme which utilizes SMVQ prediction indices to embed secret data is proposed. As a result, it improves Yang's work, which takes advantage of counts of indices in data embedding. Our approach enhances the prediction of SMVQ, resulting in a more concentrated distribution of indices. Based on the distribution of SMVQ-p indices, in the embedding process, three clusters (CL (0), CL (1) and CL (2)) are defined. Indices in CL (0) are employed to embed secret data and can support a high embedding capacity. Indices in cluster CL (1) are employed for supporting a lower bit rate. Indices in cluster CL (2) are utilized for lossless reconstruction. Experimental results indicate that lower bit rate and higher embedding efficiency are obtained after encoding and embedding processes by our approach. Furthermore, experimental results demonstrate that the proposed method outperforms other state-of-the-art reversible data hiding schemes as well.