基于三维混沌系统的通用数字图像加密与恢复算法

UNIVERSAL DIGITAL IMAGE ENCRYPTION AND RECOVERY ALGORITHM BASED ON 3D CHAOTIC SYSTEM

为有效保护数字图像的信息,提出一种新的基于三维混沌系统及其改进序列的通用加密算法。首先,根据密钥及数字图像的大小m×n,产生一个长度为m×n的三维混沌序列。然后,根据子密钥及图像类型对混沌序列进行改进,使其具有理想的伪随机特性。同时,根据子密钥及图像类型将混沌序列转换为无符号整数序列。最后,对改进后的序列进行排序以生成相应的下标序列,并据此对图像进行像素坐标置乱,同时将无符号整数序列依次与对应的像素值进行异或运算以实现像素值置换。针对恶意剪切/涂鸦攻击,同时提出一种基于邻域相邻像素特性的恢复算法。应用评价指标对加密效果与安全性进行分析。实验结果与理论分析表明,该算法运行效率较高,密钥空间巨大,具有良好的加密效果、统计特性与安全性,且抗剪切/涂鸦攻击等抗干忧能力较强。

In order to protect digital image information effectively, a new universal encryption algorithm based on 3D chaotic system and its improved sequences is proposed. Firstly, according to the key and the digital image size m × n , a 3D chaotic sequence with length of m × n is generated. Then, the chaotic sequence is improved to possess the ideal pseudo-randomness property according to the sub-key and the type of image. Also according to the sub-key and the type of image, at the same time the chaotic sequence is converted to unsigned integer se- quence. Finally, the improved sequence is sorted for generating the corresponding subscript sequence, and pixel coordinates of the image will be scrambled based on it. Meanwhile, the pixel value permutation is realised by using the XOR operation between the unsigned integer se- quence and the corresponding pixel value in turn. In allusion to the malicious shearing or painting attacks, a recovery algorithm based on neigh- bouring pixels characteristic is presented simultaneously. The effect of the encryption and security is also analysed by using the evaluation in- dex. Experimental results and theoretical analysis show that the algorithm has high running efficiency, large space of keys, good encryption effect, statistical characteristics and security, it also has strong ability in resisting the shearing or painting attacks and anti-noise as well.