Особливості шифрування тривимірних зображень

Abstract

The paper examines modern approaches to encryption of three-dimensional images and models. The use of multidimensional chaotic systems, hybrid cryptographic schemes, and deep learning methods for protecting 3D data is analyzed. The specific features of permutation and diffusion operations in polygonal meshes, point clouds, and voxel structures are described. Cryptographic strength evaluation criteria, including entropy, correlation analysis, NPCR, and UACI indicators, are considered. It is shown that the combination of chaotic models, classical cryptography, and hardware optimization ensures a high level of security for spatial data.