The chaos-based image encryption has been extensively studied for image security owing to its high efficiency. However, both the complexity of the chaotic system and the chaos-based image encryption method must be enhanced. A Lorenz-Sine coupling chaotic system with complex behavior and a novel image encryption scheme is presented in this work. By adding a nonlinear controller containing sinusoidal function to the Lorenz system, we obtain smooth chaotic attractors with a unique appearance. Analysis of the dynamical properties shows that it has a multi-symmetry stranger attractor and a large chaotic range. Based on the proposed chaotic system, a pseudorandom sequence generator is designed to generate keys with high randomness. Furthermore, an image encryption scheme using spiral-rotation and random permutation is presented. Cipher-feedback is employed to increase the avalanche effect. The encryption efficiency and the security analysis are simulated. The result shows that the correlation coefficient between the adjacent pixels of the proposed scheme achieves 10^(-4). The NPCR of the proposed scheme is larger than 0.9661 and the UACI falls within the theoretical value of UACI with a significance level of 0.05. Thus, the proposed scheme is more capable against differential attacks. Furthermore, it is more robust against noise attacks and cropping attacks.

No abstract

No abstract