A novel data-hiding and compression scheme based on block classification of SMVQ indices

This paper presents a novel, reversible, data hiding and compression scheme that hides secret data into an SMVQ-compressed image and obtains reversibility of the reconstructed SMVQ indices. The SMVQ indices are divided into index block sized 2×22×2, and four indices in each index-block are processed simultaneously. Fifteen types and four cases are defined for data embedding. Four index values in the sequence are considered together to find a corresponding case and type. Then, the pre-determined number of secret bits is embedded into each index block based on to which case or type of index block the secret bits belong. By processing a block of indices simultaneously, the proposed scheme obtains high embedding capacity, high embedding rate, and low compression rate. In addition, three different configurations, i.e., normal hiding, under hiding, and over hiding schemes, are used to achieve better improvement of the proposed scheme than several previous applications. As shown in the experimental results, the proposed scheme obtains greater efficiency than various existing schemes in terms of embedding capacity, embedding rate, and compression rate.