Novel Models of Image Permutation and Diffusion Based on Perturbed Digital Chaos

Area of Science:

• Cryptography and Information Security
• Applied Mathematics and Dynamical Systems
• Computer Vision and Image Processing

Background:

• Traditional chaos-based cryptosystems often rely on stationary chaotic dynamics, leaving them vulnerable to attacks.
• Existing image permutation and diffusion techniques in cryptosystems have known weaknesses.

Purpose of the Study:

• To propose novel models for image permutation and diffusion in chaos-based cryptosystems.
• To enhance the security and robustness of chaotic encryption against known attacks.

Main Methods:

• Introducing bit-level perturbations to chaotic maps, affecting state variables, control parameters, or both.
• Utilizing pixel coordinates, ciphered word values, and extracted state variables for perturbation.
• Implementing persistent perturbation to ensure non-stationary and image-content-dependent chaotic dynamics.

Main Results:

• The proposed models demonstrate effective image permutation and diffusion.
• Transformed images exhibit good statistical properties after a single round of encryption.
• The non-stationary nature of the chaotic map dynamics enhances resistance to cryptanalysis.

Conclusions:

• The novel perturbed chaotic map models offer a significant improvement in image encryption security.
• The proposed approach provides a robust method for secure image transmission and storage.
• Effective security is achieved with just a single round of the proposed encryption process.