查看文章

A novel Prediction Error Expansion (PEE) based reversible data hiding scheme has been proposed in this paper. A set of reference pixels are randomly distributed throughout the image. Number of these reference pixels in a portion in the image is loosely proportional to the roughness of the portion. The prediction of a non-reference pixel is carried out as the weighted median of the reference pixel values within a local neighborhood of the non-reference pixel. The data is embedded in a non-reference pixel using an adaptive embedding strategy based on an estimate of the local complexity and the estimated prediction error. The experimental results show that the proposed method outperforms some of the existing methods in the literature.