结合Hilbert曲线扫描与游程长度的图像隐写算法

A novel image steganography based on Hilbert curve scanning algorithm and run-length

针对游程长度隐写算法存在嵌入容量小的问题,提出基于Hilbert曲线扫描改进的游程长度隐写算法,该算法能增加可用于嵌入秘密信息的游程数量,也因此有效地增大了嵌入容量.新算法利用比特流中游程长度的奇偶特性来嵌入秘密信息.嵌入过程首先使用Hilbert扫描曲线对载体图像的像素进行置乱,再将置乱的载体图像分解为8个位平面,然后通过分析置乱后得到的各个位平面比特流分布来确定可用于嵌入信息的游程,最后修改在可嵌入游程的比特流中0和1交界处至多1个值来嵌入1 bit秘密信息.结果表明,基于Hilbert曲线扫描的游程长度隐写算法与已有基于游程长度隐写算法相比,前者具有较大嵌入容量,较好的视觉不可感知性.

The Run-Length steganography,referred to as RLSteg,has a problem with low embedding rate.To address this problem,based on Hilbert curve scanning algorithm,we put forward an improved RLSteg algorithm,which can increase the number of runs available for secret message embedding and thus effectively improve the embedding capacity.The presented algorithm uses the parity of run length in the bit stream to embed the secret message.The embedding process can be described as follows:the cover image is first decomposed into eight binary bitplanes,then Hilbert curve scanning algorithm is used to scramble the pixels of each bitplane.By analyzing each scrambled bitplane bit stream,the run available for secret message embedding is determined.At the edge of available run of bit stream0,1,at most one value is modified,hence1bit secret message is successfully embedded.Experimental results show that,compared with the classical RLSteg algorithm,the presented RLSteg algorithm has the advantages of larger embedding capacity and better visually imperceptible performance.