Multiple histograms based reversible data hiding by using FCM clustering
Multiple histograms based reversible data hiding by using FCM clustering
复制标题
使用 FCM 聚类基于多个直方图的可逆数据隐藏
DOI：
10.1016/j.sigpro.2019.02.013
复制
发表时间：
2019-06-01
影响因子：
4.4
通讯作者：
Shi, Yunqing
中科院分区：
文献类型：
--
作者：
Wang, Junxiang;Mao, Ningxiong;Shi, Yunqing
可逆数据隐藏算法（RDH）在多媒体版权保护和内容完整性认证中得到了广泛的应用。直方图平移（HS）作为一种典型的RDH方案，由于其高质量的隐写图像而得到广泛的研究。现有的基于HS的RDH方案大多利用预测和排序技术来构造单个锐化直方图，利用覆盖图像中的平滑区域进行数据隐藏。为了充分利用不同纹理特征的图像内容之间的相关性，近年来提出了几种基于多直方图的直方图重建算法（MH_RDH）。在本文中，聚类算法，即模糊C-均值（FCM）聚类，介绍了多直方图的建设。该方法采用FCM对覆盖载体（如预测误差）进行分类，将其分成具有相似特征的不同聚类，然后利用聚类结果构建多个直方图，实现数据的高效嵌入。实验结果表明，该方案的上级性能优于其他国家的最先进的。(C)2019爱思唯尔B. V.保留所有权利。
Reversible data hiding algorithm (RDH) has been widely used in multimedia's copyright protection and content integrity authentication. As a typical RDH scheme, histogram shifting (HS) is extensively investigated due to its high quality of stego-image. Most existing HS based RDH schemes utilize prediction and sorting techniques to build single sharp histogram, which exploit the smooth areas in cover image for data hiding. To take advantages of the correlation among image contents of different texture characteristics, several multiple histograms based RDHs (MH_RDH) are proposed recently, which resort on some rigid rules, e.g. single feature based sorting followed by uniform segmentation of sorted sequence, to construct the multiple histograms. In this paper, the clustering algorithm, i.e. Fuzzy C-means (FCM) clustering, is introduced for the construction of multiple histograms. The FCM equipped with deliberately designed features is employed to classify the cover carriers, e.g. prediction errors, into different clusters with similar traits, which are then used to build the multiple histograms for efficient data embedding. Experimental results demonstrate the superior performance of the proposed scheme over other state-of-the-art ones. (C) 2019 Elsevier B.V. All rights reserved.