Reversible data hiding scheme based on image partitioning and histogram shifting

This paper proposes a new reversible data hiding scheme based on image partitioning and histogram shifting. The existing histogram shifting reversible data hiding scheme has been presented based on finding the values of the peak point and the zero point and then moving the histogram to modify the relevant pixel values to embed information. Although this scheme has several advantages, its embedding capacity is greatly restricted, and its image quality is not very good. Aiming at this vulnerability, a new method is proposed: first dividing the image into blocks, then sorting the embedding order of the sub-images according to specific rules, and finally, embedding information according to the sorted sub-image order. The sorting rule is an ascending order based on the number of invalid pixel values that need to be modified. Experimental results show that the image embedding capacity is significantly improved with guaranteed quality. And when the amount of payload to be embedded is less than the entire payload, this scheme improves image quality by reducing unnecessary pixel value shifts.