Color Image Encryption Algorithm Based on Novel Spatiotemporal Chaotic System

Digital image security faces persistent challenges from evolving cyberattacks. In response to these challenges, a novel spatiotemporal chaotic system-Large—Parameter-Space Dynamic Non-adjacent Coupled Map Lattice (LPS-DNCML)—is proposed for image encryption. This system enhances chaotic properties by introducing dynamic coupling coefficients and an Arnold transform-based non-adjacent coupling mechanism. Superior performance is validated through comparative experiments on Kolmogorov entropy, information entropy, and mutual information. Based on the LPS-DNCML, a color image encryption algorithm is designed. The output sequence of LPSDNCML is used to construct a bidirectional square wave spiral transform to scramble the position distribution of pixels and implement a square wave channel fusion diffusion to ensuring the subtle input changes propagate globally. The new encryption algorithm is resisted the intensity of 0.2 Salt and Pepper noise and cropping attack with missing 50% part. Experimental results confirm its security, efficiency and effectiveness.