一种基于补丁算法改进的图像隐写方法简

针对经典补丁算法进行信息隐藏时嵌入容量极低的缺陷，基于图像置乱和分块的方法，提出一种改进的图像隐写算法。首先将载体图像进行恰当分块并进行充分置乱，置乱程度以分块的均方差来衡量，然后在置乱图中选择分块对进行信息嵌入。实验结果表明，该方法可将载体图像的嵌入容量由1比特扩展到数百比特，同时具有视觉不可觉察性，对JPEG压缩和噪声攻击有较好的鲁棒性，可抵御卡方、RS和SPA分析等平面域的经典隐写分析算法。

An Improved Image Steganographic Algorithm Based on Patchwork

Classical patchwork algorithm provides very low embedding capacity, so an improved steganographic algorithm based on patchwork and image scrambling is proposed. Firstly, the cover image is separated to suitable blocks, and scrambled to the best degree. The mean square error of blocks is used to evaluate the image scrambling degree. Then image block pairs are chosen for embedding. Experimental results show that the proposed algorithm can improve the embedding capacity from 1 bit to several hundreds of bits and the embedded image has good visual quality. It is also robust against JPEG compression processing and attack such as noise, and it can resist ,RS and SPA analysis.