点击次数：

论文类型：期刊论文

发表时间：2016-01-01

发表刊物：International Journal of Network Security

收录刊物：EI

卷号：18

期号：4

页面范围：750-757

ISSN号：1816353X

摘要：Prediction-error expansion (PEE) is an important reversible data hiding technique, which can hide large messages into digital media with little distortion. In this paper, we propose a nearest neighborhood pixel prediction algorithm (NNP2) for reversible data hiding algorithms based on Chinese Remainder Theorem (CRT), in which a rhombus prediction is applied in NNP2, and prediction errors, the difference between pixels and predictions, are modified to embed data. Further, CRT is utilized to adjust the modification size, thus embedding several bits into one embeddable pixel. Laplacian-like distribution of prediction errors is exploited to achieve a trade-offbetween embedding capacity and visual quality. Experimental results demonstrate that the NNP2achieves better embedding capacity with thesame stego image quality than the conventional methods.