利用差值扩展和直方图平移的可逆数字水印算法  

An Algorithm for Reversible Watermarking Based on Difference Expansion and Histogram Shifting

作　　者：王继军[1] 

机构地区：[1]广西财经学院信息与统计学院,南宁530003

出　　处：《小型微型计算机系统》2014年第5期1192-1195,共4页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61262002)资助;广西自然科学基金项目(2013GXNSFBA019274)资助;广西财经学院科研项目(2013B050)资助;国家哲学社会科学基金项目(12&ZD041)资助

摘　　要：可逆数字水印技术在遥感、医学、军事等敏感图像应用领域前景广阔,针对大多数情况下可逆水印算法需要大量附加信息的情况,提出一种附加信息最多不超过一字节的可逆水印算法.该算法主要是利用像素点灰度值之间的差值关系,先将相邻像素点灰度差值进行分类和扩展,使得一些特殊灰度差值出现空缺,扩差可能导致极少数灰度值产生溢出,可通过直方图平移得以解决;水印的嵌入以差值的大小作为判断依据,在空缺差值所留出的空间内嵌入水印,既能保证对载体图像同一像素点的修改不超过一个灰度值,又能完整提取水印,无损还原载体,同时还给出了直接分组和循环分组两种嵌入方法.理论和实验表明,算法设计简单、运行效率高、隐蔽性好、实用性强.Reversible watermarking technology has been widely used in remote-sensing, medical and military fields. Concerning most of the existing reversible watermarking algorithms need a lot of additional information, this paper presents a reversible watermarking algorithm which is no more than one byte additional information in the worst situation. This algorithm depends on the relationship of the difference between near pixel of the image. First, classifying and expanding the gray difference, which makes some special differ- ence vacancies. Difference expansion may make very few data overflow, which can be solved by histogram shifting. Then the size difference decides how to embed a watermark, and embedding method can choose direct grouping or circulation grouping. In the whole process, modification of the image is very small. Ensure that the changes of the same pixel are no more than one gray value, and the watermark and image carrier can be restored completely. Theoretical analysis and experimental results show that the algorithm is reversible and effective and secure.

关 键 词：可逆水印 差值扩展 直方图平移 信息隐藏 

分 类 号：TP309[自动化与计算机技术—计算机系统结构]