针对现有基于混沌的医学图像加密算法的鲁棒性和效率不足,提出了一种基于快速混沌置乱的鲁棒型医学图像加密算法RMIEF-CS。算法利用两个低维的混沌系统交替迭代产生混沌序列,较好地解决了由于计算机精度有限而带来的混沌收敛问题;然后利用生成的混沌序列对图像明文数据流进行第一次置乱加密,对得到的密文采用新的混沌序列进行再次置乱得到最终密文。置乱过程中引入双向密文反馈机制增加算法的安全性和鲁棒性;算法利用低维的混沌系统生成密钥,在置乱过程中无需耗时的排序操作,并适合于任何形状的图像,具有较好的时间效率和通用性。通过仿真实验验证了RMIEF-CS具有较好的加密性能,并能在密文受损情况下解密出原始医学图像的近似版本。另外,RMIEF-CS比基于均匀置乱和混沌映射的加密方法在时间效率上提高6倍左右,因而能适用于大数据量的医学图像实时保密传输。
In order to improve the robustness and efficiency of medical image encryption algorithm based on chaos, a new robust medical image encryption algorithm based on rapid chaotic scrambling named RMIEF-CS was presented. Firstly, the presented algorithm utilized two low dimensional chaotic systems to generate chaotic sequence with an alternating iterative way, and the problem of chaotic convergence due to computer precision was solved. Secondly, the data stream of plaintext image was firstly scrambled using the generated chaotic sequence, and the eiphertext was scrambled once again using a new chaotic sequence to obtain the final ciphertext image. In the second scrambling procedure, a bidirectional ciphertext feedback mechanism was used to enhance the security and robustness of RMIEF-CS. Because the proposed algorithm used the simple low chaotic system to generate key sequence, and did not need the time-consuming sort operation, it had good time efficiency and could be suitable for images with any shape. The simulation experimental results show that the presented algorithm has better encryption performance, and can decrypt the approximate image to the original medical image even if the ciphertext image has been damaged. In addition, compared with the method based on even scrambling and chaotic mapping, the time consumption of RMIEF-CS is reduced to 1/6. The presented algorithm is suitable for transmitting the medical image with large amount of data in real-time.