SCD-CHAOS: dynamic S-box and chaotic hybrid adaptive image encryption using multi-map diffusion.

The rapid growth of Internet of Things (IoT) devices has increased the demand for encryption techniques that simultaneously ensure high security strength with low computational complexity and real-time feasibility under resource constraints. The proposed SCD-CHAOS, a novel hybrid chaotic image encryption technique that combines dual chaotic dynamics based on the Tent and Henon maps with an entropy-driven adaptive dynamic S-Box mechanism. Unlike traditional chaos-based permutation-diffusion architectures which rely on static substitution structures, the proposed technique introduces entropy-guided nonlinear transformation combined with hybrid chaotic key synthesis to significantly improve confusion capability, diffusion uniformity, and key sensitivity while preserving deterministic reversibility. For the purpose of standardized benchmarking and computational efficiency in IoT environments, the input images are uniformly down-scaled to [Formula: see text] prior to encryption. Various experimental evaluations of the benchmark images demonstrate a near-ideal entropy in average of 7.9991 bits, Number of Pixels Change Rate average 99.6226%, unified average change intensity reaching average of 33.4732%, and correlation coefficients of encrypted images approaching zero in all directions, confirming strong resistance against statistical and differential attacks. Key sensitivity analysis achieves NPCR/ UACI values of up to 99.6221% and 33.4717%, respectively. The Lyapunov exponent analysis reports robust chaotic behavior with values of up to 1.4628. The detection performance shows a structural similarity index in average of 0.9994 for original vs Decrypted images, peak signal-to-noise ratio mean of 9.6250dB, and mean Square Error mean of around 6904.83. The encryption and decryption processes require only 0.00619 seconds and 0.00261 seconds for [Formula: see text] images, demonstrating real-time feasibility, making SCD-CHAOS a robust and scalable encryption technique for secure image transmission in next-generation IoT networks.