面向噪声图像的自适应高容量隐写算法

面向雷达、遥感图像等含噪图像，提出了一种利用噪声自适应嵌入的位平面加密算法。首先对不同的噪声图像利用遗传算法自适应找到最佳嵌入阈值，在噪声值较高处使用一种无须密钥的加密算法将两位秘密信息分别交叉嵌入到第一、三位平面中。然后利用BBE(binary-block embedding)算法将标记位数据压缩到第二位平面。最后在第二位平面冗余处用同样的加密算法将剩余秘密信息分别嵌入到第二、四位平面中完成信息隐藏。实验分析表明在满足高容量嵌入的条件下，与其他自适应位平面隐写算法相比有良好的隐蔽性以及较强的鲁棒性，同时能抵抗目前较好的隐写分析算法，保证了数据传送的安全性。

Adaptive high capacity steganography algorithm for noisy images

This paper proposed a bit-plane encryption algorithm using noise adaptive embedding for noisy images such as radar and remote sensing images. Firstly, the algorithm adaptively found the best embedding threshold in different noisy images using genetic algorithm, and used an encryption algorithm that did not require a secret key at higher noise values to cross-embed two pieces of secret information into the first and in the three-dimensional plane. Then it used the BBE algorithm to compress the flag bit data to the second bit plane. Finally, it used the same encryption algorithm to embed the remaining secret information into the second and fourth bit planes to complete information hiding at the second bit plane redundancy. Experimental analysis shows that under the condition of satisfying high-capacity embedding, compared with other adaptive bit-plane steganography algorithms, it has good concealment and stronger robustness, and at the same time, it can resist the current better steganalysis algorithms to ensure security of data transmission.