Advanced octree-based reversible data hiding in encrypted point clouds

This study presents an effective algorithm for reversible data hiding in encrypted point clouds, employing an advanced octree-based subdivision to significantly improve the embedding rate. By intelligently dividing the point cloud's boundary volume into distinct subspaces, each specifically adapted to the distribution of points, the octree enables accurate subspace allocation without requiring the points' positional information, thanks to its spatial organization prowess. Our algorithm advances the field of reversible data hiding in encrypted point clouds by leveraging octree subdivision and multi-MSB prediction, collaboratively enhancing the embedding rate and capacity. The algorithm skillfully adjusts the subdivision threshold, thus optimizing the subspace sizes to meet various embedding capacity needs. It also enhances the selection of pivotal reference, the subspace center, for embedding capacity calculation. The algorithm achieves an 100 % embedding rate and an average embedding capacity of 39.76 bits per point under optimal subdivision parameters, surpassing existing techniques. Comparative studies demonstrate its superior performance, with a 13.28 % increase in pure embedding capacity compared to previous methods. The algorithm guarantees the retrieval of the embedded message and the perfect restoration of the original model, facilitated by the octree's accurate point repositioning feature. These results represent a substantial advancement in reversible data hiding, promoting increased effectiveness and security for encrypted point clouds, with potential implications in multiple industries.