基于分块最佳预测的可逆数据隐藏  

Reversible data hiding based on optimal prediction of block

作　　者：熊志勇[1] 王江晴[1] 

机构地区：[1]中南民族大学计算机科学学院,湖北武汉430074

出　　处：《光电子．激光》2015年第8期1568-1574,共7页Journal of Optoelectronics·Laser

基　　金：国家自然科学基金资助项目(60975021)资助项目

摘　　要：为了提高嵌入算法的性能,提出一种基于分块最佳预测和直方图平移的灰度图像可逆数据隐藏算法。根据嵌入阈值,统计每一图像块中4种预测方案的嵌入容量,选择最大容量对应的预测方案作为该图像块的最佳预测方案。嵌入算法充分利用像素值局部相关性,自适应地选择最佳预测方案,有效提高嵌入容量。实验结果表明,本文算法在保证图像质量的同时有效提高嵌入容量,算法的整体性能比其它同类算法更优。Among the various information hiding schemes, the reversible data hiding has achieved substan- tial attention in the recent years. In order to improve the performance of embedding algorithm, this paper proposes a reversible data hiding scheme based on optimal prediction of block. The premise of this algorithm is that a histogram is constructed from the optimal prediction of block. In accordance with the priority principle of embedding capacity,the optimal prediction is chosen for every block. In the data embedding process, a histogram shifting scheme is used to embed a secret message into pixels whose prediction error is located in embedding area within the histogram. Experimental results show that the proposed al- gorithm can achieve higher embedding capacity and imperceptible distortion. Performance comparisons with other existing algorithms are also provided to demonstrate the feasibility of our proposed algorithm.

关 键 词：可逆数据隐藏 分块最佳预测 预测误差 直方图平移 

分 类 号：TP309.2[自动化与计算机技术—计算机系统结构]