A Pixel-Based Reversible Data Hiding Method Based on Obtuse Angle Prediction

Pixel-based pixel-value-ordering (PPVO) has been demonstrated to be an outstanding algorithm for reversible data hiding on digital images. High-fidelity marked images were generated from an effective pixel-by-pixel mannered prediction strategy. As an extension of PPVO, this paper proposes an novel prediction scheme, which employs the just adjacent pixel of the top-left one in a scanning block to be the current pixel. Therefore, new prediction distribution named obtuse angle prediction (OAP) is rebuilt. Experimental results declare that higher PSNR values can be achieved by the proposed OAP approach than PPVO and some other state-of-the-art methods, especially in the moderate and large payload sizes.