A smart agriculture image protection scheme based on annealing algorithm and affine transformation is optimized for S-box generated by chaos

To strengthen image security in smart agriculture, this paper presents a Two-Dimensional Super-Attractor Logistic Coupled Chaotic Model (2D-SALC). The model underwent rigorous testing against recent chaotic models, achieving a maximum Lyapunov Exponent (LE) of 13.19, a 0-1 test value of 0.9978, and a sample entropy of 2.1956, all of which passed the NIST test. Comparative analysis revealed that the 2D-SALC surpassed prior models, particularly in generating S-boxes with enhanced nonlinearity. Additionally, the integration of annealing algorithms and affine transformations further improved S-box performance, achieving a BIC-NL of 111.60, a BIC-SAC of 0.5007, a SAC of 0.5024, and a DP of 6, surpassing recent benchmarks. The model was also applied to XOR-based diffusion scrambling for image encryption, demonstrating greater key sensitivity, expanded key space, and higher information entropy. These results highlight the model's robustness and offer innovative solutions for enhancing image security in smart agriculture.