Exploiting 4D non-degenerate chaotic system and adaptive diffusion based on the finite field for 3D model encryption

Abstract With the rise of meta-universe, the application of 3D models in various fields has garnered significant attention, for they may furnish more vivid and detailed representations. However, many problems lie in securing the transmission of 3D model data, as the tampered possibility increases while in transit. Therefore, protecting 3D model data is crucial. Meanwhile, existing chaotic systems have many drawbacks such as degeneracy, predictability, and initial value insensitivity, resulting in a high risk of attacks on 3D model encryption algorithms. This study proposes a 3D model encryption algorithm using a 4D non-degenerate chaotic system and adaptive diffusion based on the finite field. Firstly, a 4-dimensional (4D) non-degenerate chaotic system is designed for generating pseudorandom sequences. Compared to the classical 4D chaotic systems, our proposed chaotic system exhibits better chaotic performance in several aspects, such as Lyapunov exponents, NIST test, and initial value sensitivity. Further, the 3D data format of the 3D model is transformed into a 2D data format. Then, a rotation misalignment across rows and columns (RMARC) is presented to confuse the 2D coordinate data format and break the high correlation of the 2D coordinate values. Finally, the scrambled data is diffused within a finite field using the chaotic sequence to get the diffused 3D model. In addition, we have done many experiments to verify the encryption efficiency, security performance, and encryption quality. The experimental results show that the 3D encryption system performs well in these aspects.