Hybrid trigonometric-logarithmic chaotic encryption scheme for UAV captured images using dynamic S-box confusion and non-linear diffusion

Unmanned aerial vehicles (UAVs) are becoming valuable platforms for data collection in various application areas, such as agriculture, environmental monitoring, surveillance and infrastructure inspection. In the proposed framework, the UAV is used for real-time image acquisition, while the captured images are securely transmitted to a ground processing system where the proposed HTLC-based encryption and decryption operations are executed. To maintain data privacy and security, this research presents a novel Hybrid Trigonometric-Logarithmic Chaotic (HTLC) encryption approach for UAV-captured images. The proposed HTLC map demonstrates strong chaotic behavior, higher positive Lyapunov exponent, a wide chaotic range and successfully satisfies the National Institute of Standards and Technology (NIST) statistical randomness tests. The designed encryption scheme integrates four distinct stages: Key and seed initialization, Chaotic sequence generation, Dynamic S-Box-Controlled Scrambling (SCS), and Non-Linear Pixel Mixing (NLPM) diffusion. In the scrambling phase, two chaotic S-boxes and a reverse index-based permutation strategy are employed to effectively reduce spatial pixel correlation. The diffusion phase employs key-dependent XOR operations, modular arithmetic, integrated with nonlinear mixing, bit-level rotations, and chained feedback mechanisms to enhance diffusion diversity. Experimental results demonstrate that the proposed scheme achieves high security performance with information entropy values close to 7.999, correlation coefficients near zero, Number of Pixels Change Rate (NPCR) values around 99.62% and Unified Average Changing Intensity (UACI) values up to 33.44% The encrypted images also exhibit a uniform histogram distribution and accurate decryption capability, confirming the robustness and effectiveness of the proposed framework for secure UAV image communication.