Research on Information Hiding Algorithm Based on Local Fusion of 3D Model

Aiming at the problem that the single-carrier 3D model has a weak ability to resist heterogeneous compression attacks, an information hiding algorithm based on the local feature fusion of the multi-carrier 3D model is proposed. First, the PCA (principal components analysis) regression method was used to estimate the normal carrier of the surface curvature, and the feature points of the vector set of 3D model were extracted from the surface curvature; Secondly, the carrier set was prefused with spindle correction, and then the distance threshold between feature points was set by ICP (Iterative Closest Point) nearest point iterative algorithm. After the fusion state of the carrier was obtained, the local fusion block was extracted by difference calculation with the original carrier; Thirdly, Arnold is used to scramble the secret information to get the binary sequence with the optimal matching degree with the carrier; Finally, meshed local fusion blocks, whose structural complexity is defined by the perimeter of the grid, embed binary sequences into vertices with high structural complexity. The algorithm hides the binary sequences with optimal matching degree with the carrier in the vertices with high structural complexity in the local fusion block, the invisibility and robustness of the algorithm are improved greatly. It can be seen from the experimental results that the proposed algorithm has high resistance to non-uniform compression and other attacks, indicating that its invisibility and robustness are high.