A Secure Communication System Based on Self-organizing Patterns

This paper proposes a secure steganographic communication algorithm based on the evolution of self-organizing patterns. The presented algorithm is a modification of a secure steganographic scheme, presented in our previous work [1]. Algorithm is based on the formation of self-organizing patterns in a Beddington-deAngelis-type predator-prey model with selfdiffusion. Computational experiments show that the generation of interpretable target patterns cannot be considered as a safe encoding of secret visual information because the target pattern becomes interpretable only when the cover image (initial distribution of preys) leaks the secret to a naked eye. Therefore, we propose an alternative approach when the cover image represents the self-organizing pattern which has evolved from initial states perturbed using the dot-skeleton representation of the secret image. Such visual communication technique protects both the secret image and communicating parties.