Block-based reversible data hiding with multi-round estimation and difference alteration.

In this paper, we propose a new algorithm in reversible data hiding, with the characteristics of ease of implementation, based on the alteration of difference values relating to original image. Three major categories can make reversible data hiding possible, one is to modify the histogram of original images, the second is to alter the difference between neighboring pixels, and the third is to apply prediction method to modify the difference between original and predicted pixels. By integrating the advantages from the these categories, and by extending the flexibility to acquire the difference values for data hiding, we employ the block-based, multi-round prediction to look for enhanced performances. Simulations with conventional and medical images have presented the superiority with our algorithm over existing ones in literature. With our algorithm, enhanced amount of capacity, in addition to the better image quality, and most important of all, the reversibility of algorithm, can be reached. It is also famous for its ease of implementation for medical applications in hospitals. [ABSTRACT FROM AUTHOR]