基于综合压缩策略的加密域可逆信息隐藏算法

Reversible Data Hiding in Encrypted Images Based on Comprehensive Compression Strategy

摘要: 加密图像可逆数据隐藏(Reversible Data Hiding in Encrypted Images, RDHEI)属于多媒体信息安全技术重要研究内容。该技术不仅能够通过加密的载体图像传输秘密数据，还能在不损失信息的情况下恢复载体图像，并准确提取秘密数据。加密后腾出空间(Vacating Room After Encryption, VRAE)是RDHEI的一个常用框架，但其嵌入容量较低。本文提出了一种基于子块的加密方法，即在同一块内使用相同的密钥进行逐位流加密，以保留块内像素间的相关性。为增强安全性，使用加密密钥K_e置乱图像子块的排列顺序。在数据隐藏过程中，通过分析子块内像素连续相同的高位平面数量，将该数量作为子块的块标签。引入参数T，通过块标签与参数T进行比较，为不同子块选择最优压缩策略。具体来说，对于块标签大于T的子块，采用相同位平面压缩策略，并以熵编码形式进行标记；对于块标签小于T的子块，采用保留差值的压缩策略。实验结果表明，在保证安全性和可逆性的前提下，提出方法在嵌入性能上实现了显著的优化效果。

Abstract: Reversible Data Hiding in Encrypted Images (RDHEI) is an important research topic in multimedia information security technology. This technology can not only transmit secret data through encrypted carrier images, but also recover carrier images without losing information and accurately extract secret data. Vacating room after encryption (VRAE) is a common framework for RDHEI, but its embedding capacity is low. In this paper, we employ a block-based stream encryption method, i.e., bit-by-bit stream encryption using the same key within the same block to preserve the correlation between the pixels within the block. To enhance security, the ordering of image blocks is scrambled using the encryption key K_e. During the data hiding process, the number of consecutively identical high planes of pixels within a sub-block is calculated using that number as the block label of the sub-block. This paper introduces parameter T and compares it with block labels to select the optimal compression strategy for different sub blocks. Specifically, for blocks with labels greater than T, the same bit plane compression strategy is used and labeled in entropy encoding form; For blocks with labels less than T, a compression strategy that preserves the difference value is adopted. The experimental results show that the proposed method achieves significant optimization effects in embedding performance while ensuring safety and reversibility.