基于分块最佳预测的可逆数据隐藏

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。

Reversible data hiding based on optimal prediction of block

Among the various information hiding schemes,the reversible data hiding has achieved substantial attention in the recent years.In order to improve the performance o f embedding algorithm,this paper proposes a reversible data hiding scheme based on optimal p rediction of block.The premise of this algorithm is that a histogram is constructed from the optimal pr ediction of block. In accordance with the priority principle of embedding capacity,the optim al 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 algorithm can achieve higher embeddi ng capacity and imperceptible distortion.Performance comparisons with other existing algori thms are also provided to demonstrate the feasibility of our proposed algorithm.