An Improved Image Encryption Algorithm Based on Chaotic Maps

This paper proposes a 3-D Lorenz system based image encryption scheme, which improves the security and performance of the encryption system over conventional one dimension chaos based ones. The balance and correlation properties of the analog sequences generated by Lorenz system are analyzed and a pre-process method is proposed, which makes the key stream have good statistical properties. Three initial parameters of Lorenz system are combined together as one key, which greatly enlarges the key space under precision restricted condition. The encryption speed is also improved since the position permutation and grayscale substitution of a pixel can be done by one iteration operation. Besides, the encryption system is strong against adaptive parameter synchronous attack since it has more complicated structure than one dimension ones. The experimental results indicate that the algorithm is more secure and efficient than conventional ones.