Cross-channel image chain scrambling encryption algorithm using two memristor-based twin multi-scroll hyperchaotic systems

This paper proposes a cross-channel image chain scrambling encryption algorithm based on two 5D memristive twin multi-scroll hyperchaotic systems. Firstly, two flux-controlled memristor models are introduced based on the simplification of Chua's system, and two 5D memristive twin multi-scroll hyperchaotic systems are designed by inputting different flux variables. The twin 5D memristive multi-scroll hyperchaotic systems exhibit the characteristics of variable scroll numbers and different topological structures. Secondly, a fractal-like cross-channel encryption algorithm is proposed, which combines cross-channel scrambling of sub-image and Zigzag transformation. Based on comprehensive experimental analysis, the results reveal that the key space is determined to be 2 525 , the information entropy of the ciphertext images across three channels approaches 7.9994, and the pixel correlation is reduced to nearly zero. Furthermore, NPCR and UACI exhibit values close to their theoretical benchmarks of 99.6094 % and 33.4635 %, respectively, highlighting the superior performance of the proposed encryption algorithm in terms of robustness and security. • Two new memristive twin multi-scroll hyperchaotic systems are designed, which have different topologies. • Two key-acquiring methods are proposed: SHA-256 obtains key 1 from image 1; image 2 combined with key 1 gets key 2. • The proposal of sub-image cross-channel disturbances: this method uses chained execution. • The proposal of a fractal-like cross-channel image scrambling encryption strategy combined with Zigzag transformation.