A Reversible Data Embedding Scheme Based on Hash Functions for VQ Index Tables

A data hiding scheme with flexible hiding capacity applied in vector quantization (VQ) index tables was proposed by Chang et al. However the minimum bit rate of their scheme is 0.552. To improve the bit rate of Chang et al´s method, this paper proposes a novel and efficient reversible data embedding scheme for VQ index tables. The proposed method exploits characteristics of VQ indices and uses two hush functions to embed the secret data, which is able to make sure that the original indices can be restored to reconstruct the VQ compressed image. The experimental results show that the proposed scheme can achieve the best bit rate and have the satisfactory hiding capacity for all test images.