基于仿射和复合混沌的图像加密新算法

An Encryption Algorithm for Image Based on Affine and Composed Chaos

随着多媒体的广泛应用,数字图像的安全性变得越来越重要.针对数字图像的特点,基于三维仿射变换和混沌,提出一种新的空域加密算法.先置乱像素的位置并根据像素坐标混合像素的值,然后按行交替进行非线性的扩散、代换,如此迭代至少3轮.代换时用中间结果扰动耦合的多个混沌系统以进行自适应的加密,置乱参数由混沌系统自动生成,置乱操作可以直接作用于任意宽高比的图像,不需要进行预处理.稍加改动算法使之处理对象为频域量化后系数,便可转换为频域加密算法.理论分析表明:算法密钥空间巨大,可抵御穷举攻击;明密文映射关系复杂,可有效地抵御选择明文攻击;算法符合模块化设计思想,采用的混沌系统形式简单,易于并行实现.实验结果表明:算法加密效果好,敏感性强,符合密码学中的混淆与扩散原则,安全性高.

With the popular application of multimedia, the security of digital image becomes more and more important. According to the feature of digital image and on the basis of three-dimension affine transformation and chaos, a novel spatial domain encryption algorithm is proposed. Firstly, it scrambles pixel position and confuses pixel value according to the corresponding coordination. Secondly, it takes a series of nonlinear diffusion and substitution in turn for all lines. The algorithm proceeds with the above two steps for at least 3 times, and it can be conveniently converted to the frequency domain algorithm while replacing the processed data in the spatial domain with quantized coefficients in the frequency domain. In the process of substitution, pixel value is introduced to perturb multiple chaos systems that are coupled together for self-adaptive encryption. In the encryption process, the scrambling parameters are generated by chaos systems automatically, and the scrambling function is compatible with images at any ratio of length to width without any preprocessing. Theoretical analysis shows that the algorithm has huge key space to defend against violent attack, the mapping relation between the plaintext and the ciphertext is complex enough to resist chosen plaintext attack efficiently, and the algorithm using simple chaos systems is designed modularly in order to be realized parallelly conveniently. Experimental results show that the algorithm takes good encryption result, gets strong sensitivity, conforms to confusion and diffusion principles in cryptography, and achieves high security.