Reversible Data Hiding by Median-Preserving Histogram Modification for Image Contrast Enhancement

A new image reversible data hiding algorithm is proposed in this paper based on histogram modification. Instead of trying to keep the image PSNR value high after data hiding, the proposed algorithm aims at enhancing the image contrast for better visual effect. To make the hidden data blindly extractable, the median point in the histogram is preserved after modifying two selected bins in a special way. Multiple pairs of histogram bins are further selected to be modified in sequence to achieve the effect of contrast enhancement. To avoid overflow and underflow, pre-process is performed on the pixel values and a location map is generated. By embedding the side information along with the message bits, the original image can be completely recovered without additional information. The proposed algorithm was applied to two sets of test images and the experimental results have shown that the better effect of contrast enhancement can be achieved than the previous works. In addition to reversibility, the evaluation results show that the visual quality of test images has been preserved after data hiding, even better than three specific MATLAB functions.