An Adaptive Image Encryption Scheme Based on Iterated Function System

Abstract

A novel 2D dynamical system is derived from the iterated function system which is used to generate the Sierpinski carpet-like fractal images. Experimental results show that the derived system has better chaotic performances. Therefore we can apply the derived chaotic system to generate pseudo-random sequences to design image encryption schemes. To achieve better ability of resisting chosen-plaintext and known-plaintext attacks, we design an adaptive image encryption scheme with a permutation-diffusion structure. In the permutation stage, we generate plain-image content dependent pseudo-random sequence to change the pixels positions. Consequently, different plain-images produce distinct features for the diffusion stage. The performance and security analysis of the proposed image encryption scheme have been carried out, including histogram analysis, correlation analysis, entropy analysis, and differential attack analysis, etc. The experimental simulations demonstrate that the proposed image encryption scheme is highly secure and reliable for practical image encryption.