摘要

In this paper, we propose an adaptive reversible steganographic scheme based on the just noticeable distortion (JND). First, the JND value of each cover pixel is calculated using the frequency model of the human visual system (HVS). Then, the prediction value of the cover pixel is acquired by anisotropic interpolation, and also the pixel distribution characteristic is estimated. Finally, whether the cover pixel is embeddable or not is adaptively determined according to the relationship between the prediction error and the JND value. The embedding procedure is based on modifying the prediction error of each cover pixel, and the visual degradation caused by embedding is imperceptible due to the control of JND. Experimental results demonstrate that the proposed scheme provides a greater embedding rate and higher quality of stego image than other methods that have been reported recently.