Secure steganographic communication algorithm based on self-organizing patterns.

A secure steganographic communication algorithm based on patterns evolving in a Beddington-de Angelis-type predator-prey model with self- and cross-diffusion is proposed in this paper. Small perturbations of initial states of the system around the state of equilibrium result in the evolution of self-organizing patterns. Small differences between initial perturbations result in slight differences also in the evolving patterns. It is shown that the generation of interpretable target patterns cannot be considered as a secure mean of communication because contours of the secret image can be retrieved from the cover image using statistical techniques if only it represents small perturbations of the initial states of the system. An alternative approach when the cover image represents the self-organizing pattern that has evolved from initial states perturbed using the dot-skeleton representation of the secret image can be considered as a safe visual communication technique protecting both the secret image and communicating parties.