Dynamic Reversible Data Hiding for Edge Contrast Enhancement of Medical Image

Abstract

Reversible data hiding (RDH) for medical image contrast enhancement is designed to effectively improve the quality of medical images to help doctors make correct diagnosis, while addressing issues of privacy protection and image content integrity. In this paper, we propose a new RDH method for medical image contrast enhancement. To enhance the edge contour of medical image, we employ the superpixel segmentation to identify region of interest (ROI), and then improve the region contrast to facilitate the diagnosis. A new histogram modification is proposed to achieve a local histogram equalization effect. Two adjacent bins with the largest difference in number are selected for expansion, in order to spread the histogram evenly as much as possible. In addition, the histogram modification is adaptive to the expansion bins by using the multiple modification manner, and can spread out the highly populated bins more evenly. Experimental results verify that, compared with the existing typical methods, the proposed method can better improve the medical image quality after data embedding in terms of contrast.