Reversible Data Hiding Method using Surrounding Context Predictors

Abstract

Reversible data hiding based on prediction error expansion method continues to attract significant attention in recent years due to its high embedding capacity and good image quality. Furthermore, the prediction method is widely used and is especially effective in reversible information hiding on cryptographic images. One of the factors that affect the effectiveness of a reversible data hiding scheme is the predictor. In this paper, we propose a new predictor in which pixels that are more correlated in the prediction context will have more priority. This predictor will help exploit the correlation between pixels more effectively and have smaller prediction errors. Therefore, applying this predictor to the reversible data hiding scheme will give better capacity. Experimental results also indicate that the proposed method yields higher embedding capacity compared to other related schemes.