嵌入
像素
信息隐藏
直方图
块(置换群论)
冗余(工程)
算法
价值(数学)
计算机科学
数学
图像(数学)
人工智能
统计
组合数学
操作系统
作者
Fei Peng,Xiaolong Li,Bin Yang
标识
DOI:10.1016/j.dsp.2013.11.002
摘要
This work extends a recently proposed reversible data hiding (RDH) scheme of Li et al. which is based on pixel-value-ordering and prediction-error expansion. In Li et al.ʼs method, the maximum and minimum of a pixel block are predicted and modified to embed data. The pixel value order of each block is unchanged after data embedding and the property guarantees the reversibility. In this work, instead of the difference between the maximum and second largest value of a block (or, the minimum and second smallest value of a block) considered in Li et al.ʼs method, new differences are computed and new histogram-modification-strategy is utilized. Take the maximum for example, the new difference is defined considering the pixel locations of the maximum and second largest value. In this way, the blocks where the maximum equals to the second largest value can be exploited to embed data while these blocks suitable for RDH are not utilized in Li et al.ʼs work. This can better exploit image redundancy and achieve a superior embedding performance. Extensive experiments verify that the proposed method outperforms Li et al.ʼs and some other state-of-the-art works.
科研通智能强力驱动
Strongly Powered by AbleSci AI