Image encryption algorithm using 3-D non-equilateral Arnold transform combined with chaos matrix
Zerui Wu, Zhenlong Man, Ying Long Zhou · 2025
The arrival of the big data era has brought the threat of user privacy data leakage while promoting the development of society, in order to further safeguard the security of image information in the information age, this paper proposes an image encryption algorithm using three-dimensional non-equilateral Arnold transform combined with chaotic matrices, in the initial stage, the plain-image is split in three dimensions, and the image is subjected to a 3-D non-equilateral Arnold transform, thereby resulting in its scrambled state. Next, the sequences generated by the 4-D hyperchaotic system are reshaped into a diffusion matrix for performing diffusion, and the sequences generated by the cellular automata are split and processed at both ends using a sliding window with a window size of 8 and a step of 8. A portion of the sequences are reshaped as a second diffusion matrix with a size equal to the plain-image, and the other portion of the 1-D sequences has the effect of creating a scramble effect. Finally, the 3-parameter fractal matrix is utilized to diffuse the image to finally obtain the cipher-image. In the meantime, the 4-D hyperchaotic system and Hénon map used as generate parameters and diffusion sequences required by the encryption algorithm. The experimental analysis of six grayscale images “Peppers”, “Lake”, “Cameraman”, “Boat”, “Jelly Beans” and “House” verifies the feasibility of the algorithm and the strong encryption effect.