Reversible Data Hiding Based on a Sorted VQ Index Table

Abstract

Recently, a reversible VQ-based data embedding scheme which emphasizes that the original VQ compressed codes can be recovered after data extracting was presented by Chang et al. In their scheme, a sorted VQ codebook is divided into 2B13 clusters and indexes located in the front onethird clusters are used to embed secret data, where B denotes the size of secret data embedded into each VQ index. In this paper, a new reversible VQ-based hiding scheme is proposed. In our scheme, a sorted VQ codebook is divided into 2B clusters and half of clusters are used to embed secret data. Strategies of priorities, indicators, and index exchanging are proposed to improve our scheme further. Under the same sorted VQ codebook, experimental results demonstrate that our data hiding algorithm has higher capacity and better compression rate.