Large Quantities of Acoustic Multibeam Bathymetric Point Clouds: Organizing Method for Efficient Storage and Retrieval
2025
Xianhai Bu | Shuaibing Dou | Jianxing Zhang | Tianyu Yun | Yabing Zhu | Yi Huang | Xiaodong Cui
To efficiently organize large quantities of acoustic multibeam bathymetric point clouds, this paper proposes an improved oriented quadtree-based method for establishing a data indexing structure stored on a hard disk. First, the spatial characteristics of the multibeam swath data are integrated into the traditional quadtree structure, resulting in an oriented quadtree for data organization. Then, the primary orientation of the root node&rsquo:s bounding box, which reflects the main orientation of the swath, is consistently applied to all child nodes, eliminating the need to calculate the orientation for each individual child node by the conventional oriented quadtree. Finally, index files containing the point cloud offset, oriented bounding box, and child node information for root, child, and leaf nodes are designed and stored in external storage. Experimental results indicate that, in terms of tree construction time, although the traditional quadtree reduces time consumption by approximately 50% compared to the improved oriented quadtree, the improved oriented quadtree still achieves a 70% reduction in time consumption compared to the conventional oriented quadtree. Regarding point cloud retrieval, within the same retrieval range, the improved oriented quadtree achieves similar average retrieval times as the conventional oriented quadtree but reduces the maximum time consumption by approximately 20.83% compared to the traditional quadtree. Furthermore, by storing the constructed index in binary format on external storage, the space occupancy was reduced by 50%. The approach effectively organizes acoustic multibeam bathymetric point clouds, providing valuable insights for enhancing point cloud retrieval efficiency and reducing data memory usage.
Показать больше [+] Меньше [-]Ключевые слова АГРОВОК
Библиографическая информация
Эту запись предоставил Multidisciplinary Digital Publishing Institute