提出了一种新的基于图像复杂度的隐写算法。该算法充分利用了复杂度高的图像, 包含更多的信息, 因而能承载更多秘密信息的特点, 将图像划分为3×3的图像块, 借助图像块的信息熵和对比度描述图像块的复杂度, 根据各块复杂度统计值确定阈值, 以判定各块复杂度高低, 并针对不同复杂度的图像块进行不同强度的嵌入, 实现了基于图像复杂度的隐写算法。性能分析与实验对比表明, 在相同的嵌入率下, 该算法具有较高峰值信噪比和较强的统计安全性。

A new adaptive steganographic algorithm based on image complexity in space domain was proposed. The algorithm takes full advantage of the images feature that the image with high complexity contains more information and thus can load more secret message. In order to measure the complexity of image blocks in a better way, the algorithm divides the cover image into 3×3 pixel blocks, and then measures the complexity of blocks by computing information entropy and contrast. After that, the algorithm analyzes the statistics of all the blockscomplexity to get the threshold for evaluating the complexity of every block, and then chooses steganographic algorithms with different strength at the different complexity of blocks. The performance analysis and comparison experiments showed that in the same embedded rate, the proposed algorithm has higher RPSN and stronger statistical security in the same embedded rate.