Color image encryption algorithm based on a novel five-dimensional chaotic system
Xiaojun Tong, Liming Cheng, Yihui Wang
Abstract
Abstract Despite significant progress in multimedia security, current chaotic image encryption systems remain challenged by insufficient complexity, brute-force attack vulnerability, and statistical analysis susceptibility. This paper introduces an innovative image encryption algorithm leveraging a five-dimensional (5D) chaotic system. First, we construct a 5D chaotic system which exhibits highly nonlinear behavior, rigorously validating its enhanced randomness and dynamic properties through Lyapunov exponent analysis, Poincaré cross-sections, phase diagrams, as well as approximate and sample entropy tests. Second, the conventional encryption structure is refined by introducing a position cascade permutation strategy to achieve nonlinear pixel confusion, while a diffusion mechanism is implemented using cat mapping. Comprehensive evaluations demonstrate the algorithm’s superiority over existing solutions in metrics including information entropy, adjacent pixel correlation, NPCR (Number of Pixel Change Rate), and UACI (Unified Average Changing Intensity). Robustness tests against salt-and-pepper noise and clipping attacks further confirm its exceptional security. Notably, the proposed method also achieves faster encryption speeds than comparable algorithms.