Reversible data hiding based on histogram shifting in wavelet domain

摘要

A new reversible data hiding algorithm with better performance of robustness is proposed in this letter. In the scheme, the original image is firstly divided into some no over-lapping sub-blocks, then Inter Waver Transform (IWT) is applied to every image sub-block, the resulting low-frequency approximate sub-band is used to hide data based on the statistical features of image, two threshold parameters can be adaptively selected according to the peak-signal to noise rate (PSNR), such that watermarked image has better visual effect, the novelty of the algorithm lies in that a new histogram is constructed and the histogram is divided into two parts, which is used to hide data by using histogram shifting. Experimental results show that the proposed scheme is reversible, and also has better robustness to JPPEG compression attack; compared with relative literatures, the proposed algorithm also has the advantage of large capacity of data hiding and higher PSNR.