提出了一种基于遥感影像融合的不同权限机密信息隐藏盲算法,算法利用奇偶嵌入法和JPEG量化表,根据用户的不同权限进行融合影像中机密信息的不同程度的隐藏,且在提取机密信息和恢复原始融合影像时,不需要原始遥感影像,是一种盲算法。实验结果表明:算法具有强的透明性,大的信息隐藏量,并对各种常见的影像处理具有较强的鲁棒性。此外,算法对不同程度隐藏机密信息的遥感影像的各种应用均没有影响。

Remote sensing image fusion is an effective means to realize abundant presentation of enormous image data from various origins. However in the fused remote sensing image, there may be some secrete information such as military bases that cann't be opened to all users. Just those authorized users can completely and distinctly obtain the secrete information; partly-authorized users can only faintly see those secrete; but the unauthorized users cann't see it. In this paper, we proposed a blind algorithm for secrete information hiding based on remote sensing image fusion and it can achieve hiding to different degrees the secrete information in the fused remote sensing image according to different popedoms of different users. Firstly, we fused the remote sensing images from various origins, chose, segmented and extracted the secrete information in the fused image and according to different users' popedoms produced the disguised fused images by filtering and pixel complementation in which there presented the secrete information to different degrees. Then we used for reference the idea of information hiding and utilized the algorithm of odd-even embedding and JPEG standard quantization table to embed the secrete sub-image imperceptibly into the disguised fused image to produce the disguised one in which there hid secrete sub-image. While embedding, we exploited JPEG standard quantizaton table to define the quantization step which further enhanced the robustness of the algorithm against the attack of JPEG lossy compression. In addition, during the course of extracting the secrete information and resuming the fused image, this algorithm didn't need the original remote sensing images and was a blind one. And experimental results show that this algorithm is not only quite transparent and has a good effect for large amount of secrete information hiding, but also has a strong robustness against such familiar image processes as JPEG lossy compression, noise adding, scaling, filtering, rotation and cropping. Furthermore this algorithm has no influence on various applications of the disguised fused images with the hidden secrete information to different degrees, such as edge detection and image classification.