A novel image protection cryptosystem with only permutation stage: multi-shuffling process

Shih-Yu Li,Benalcázar Hernández Miguel Angel

DOI: https://doi.org/10.1007/s00500-023-07970-y

IF: 3.732

2023-03-13

Soft Computing

Abstract:In this paper, a novel cryptosystem with just a multi-shuffling process which is highly interactive to differentiate initial conditions is developed. The proposed system replaces the fundamental and well known in cryptography, confusion–diffusion scheme, which uses two levels with just one confusion block. This algorithm is designed for protecting all personal or commercial data in real-time applications such as monitoring and transmission of multimedia information. The unique confusion stage has three-level shuffling processes, consisting of the flipping procedure, block shuffling, and permutation in a bit level. All of them were designed to drastically change data from a plain image to a cipher image. This not only destroys the relationship of adjacent pixels but also destroys patterns in the encryption process. The flipping procedure tries to combine parts of the image's top with the bottom and the right with the left. The selected columns and rows will change completely after this procedure due to the complement of each pixel being written during the process. The last levels use an automatically sliding window combined with a logistic sine wave and logistic map to drive the iteration of the cryptosystem and shuffle the position of pixels of the plain image as well as giving new statistical features to the image at the same time. The proposed algorithm was designed to be easily implemented by practitioners. Additionally, the main goal of the image encryption is to achieve an uncertain image that cannot be cracked. The results show that this novel cryptosystem with only the shuffling process is well performed and feasible to encrypt images with a high security level. Furthermore, the experimental analysis demonstrates the effectiveness and feasibility of the proposed encryption algorithm after being tested using the benchmark “Lena” image, "The Wonder woman" image, and the “Bruce Lee” image, the latter of which is completely different to the first one, statistically speaking.

computer science, artificial intelligence, interdisciplinary applications