Lorenz, Rossler, and Chan systems for key generation and pixel selection in chaotic image encryption
Raghad Abed Sahun, Aws Hamad, Lamis Hamood Al-Saadi, Ameer Jawad · Journal of university of Anbar for pure science · 2025
The increasing need for secure multimedia communication has driven the development of more robust encryption schemes. This paper introduces a novel chaotic image encryption algorithm that uniquely integrates three different chaotic systems, including Lorenz, Rossler, and Chan, to generate a highly unpredictable, non-sequential encryption process. Unlike prior dual-system approaches, the proposed method assigns distinct roles to each chaotic system for key generation, pixel selection, and indexing, thereby enhancing both confusion and diffusion properties while maintaining low computational overhead. The three proposed chaotic systems, Lorenz, Rossler, and Chan, introduce a high level of randomness in the encryption process. Lorenz system generates chaotic random integer numbers (CRINs) as encryption keys to ensure unpredictability. Rossler system randomly selects nonsequential pixel indices to enhance diffusion. Chan system applies an XOR operation between the selected image pixels and the generated chaotic keys and encrypt image with highly resistant to decryption attempts. The Diffie-Hellman key exchange protocol was also incorporated to share encryption keys between parties securely. Multiple image quality assessment metrics were used to evaluate the proposed algorithm’s effectiveness. The results indicated that the encrypted images exhibited significant distortion with a Mean Squared Error (MSE) of 9254.5 and a Peak Signalto-Noise Ratio (PSNR) of 8.4673 dB, indicating strong encryption. The Structural Similarity Index Measure (SSIM) was very low (0.000604), reflecting high image scrambling. The encrypted image’s entropy was 7.9997, nearing the ideal value of 8, ensuring maximum randomness. The decryption process accurately reconstructed the original image with an SSIM of 1 and zero MSE, with encryption and decryption times of 0.409s and 0.369s, respectively. The proposed encryption algorithm offers an effective solution for protecting digital images in public communication channels. Integrating chaotic systems with a structured encryption framework shows strong resistance to modern cryptographic threats. This approach is particularly valuable in high-security areas such as military operations, medical data transfer, and confidential communications.