Lossless Information Hiding in the VQ Index Table

Abstract

In this paper, a reversible data hiding scheme based on a Vector Quantization (VQ) index table is proposed. To satisfy different needs of users, our proposed scheme is designed with flexibility in adjusting hiding capacity and compression rate. First, the codebook is rearranged referring to the index occurrence frequency in the VQ index table. After that, the codewords in the newly generated codebook are clustered into a number of groups for the usage of embedding secret digits in our proposed scheme. Note that, the decision of group numbers determines the capability of hiding capacity and compression rate of a to-be-embed image in our proposed scheme. Based on the well-defined embedding strategies, the experimental results show that the performance of our proposed scheme outperforms the scheme proposed by Yang and Lin in 2009.