Reversible Data Hiding With Multi-Round Prediction and Subsampling Concepts

Reversible data hiding has attracted attention in recent years due to the capability of reversibility and the ease of implementation. With the consideration of manipulating difference values corresponding to original image, more amount of secret data can be embedded to compare to conventional schemes. By taking inherent characteristics of original image into account, subsampling techniques can be applied for reversible data hiding with a variety of combinations to enhance security. With the execution by taking the reverse order, both original and secret information can be extracted to guarantee the reversibility. Simulation results present the effectiveness of proposed method.