基于位平面压缩的密文医学图像可逆信息隐藏算法

针对目前医学图像可逆信息隐藏算法嵌入容量小、需要对图像进行感兴趣区域（ROI）划分、接收方操作不灵活等缺点，结合医学图像特点，提出了一种基于位平面压缩的可分离式密文域信息隐藏算法。首先，将256级灰度医学图像分解成8个位平面，压缩高4个位平面，用峰值点像素值填充压缩后的空间，重构图像；然后，对重构的图像头部、中部、尾部分别加密；最后，在尾部根据嵌入密钥选取位置，通过直方图平移算法嵌入信息，接收方可根据密钥持有情况实现信息提取与图像恢复的可分离操作。实验结果表明，通过压缩图像预留空间来存放信息避免了辅助信息的传输，能有效提高嵌入容量，同时具有较高的安全性。

Reversible data hiding in encrypted medical images based on bit plane compression

The present reversible data hiding algorithms in encrypted medical image have low embedding capacity and need to partition Regions Of Interest (ROI), the receiver operation is not flexible. Combining with the characteristics of medical images, a novel separable reversible data hiding method in encrypted medical images was proposed. Firstly, the medical image with 256 gray levels was decomposed into 8 bit planes, and the highest four bit planes were compressed, which left some space to fill with the peak pixel values. Then the image was reconstructed. The head, middle, tail of the reconstructed image were encrypted respectively. After that, information was embedded in the encrypted image by histogram shifting, and the position was chosen by the data hiding key in the tail. For the receiver, information extraction and image restoration can be separated operations according to keys. The experimental results demonstrate that information can be stored in the compressed image to avoid auxiliary information transmission, which effectively improves the embedding capacity with higher security.