刘泽邦, 陈荦, 杨岸然, 马梦宇, 曹竞之. 支持大规模地理矢量数据实时可视化的高效索引技术[J]. 武汉大学学报 ( 信息科学版), 2023, 48(9): 1512-1521. DOI: 10.13203/j.whugis20210445
引用本文: 刘泽邦, 陈荦, 杨岸然, 马梦宇, 曹竞之. 支持大规模地理矢量数据实时可视化的高效索引技术[J]. 武汉大学学报 ( 信息科学版), 2023, 48(9): 1512-1521. DOI: 10.13203/j.whugis20210445
LIU Zebang, CHEN Luo, YANG Anran, MA Mengyu, CAO Jingzhi. Efficient Indexing Technology for Real‐Time Visualization of Large‐Scale Geographic Vector Data[J]. Geomatics and Information Science of Wuhan University, 2023, 48(9): 1512-1521. DOI: 10.13203/j.whugis20210445
Citation: LIU Zebang, CHEN Luo, YANG Anran, MA Mengyu, CAO Jingzhi. Efficient Indexing Technology for Real‐Time Visualization of Large‐Scale Geographic Vector Data[J]. Geomatics and Information Science of Wuhan University, 2023, 48(9): 1512-1521. DOI: 10.13203/j.whugis20210445

支持大规模地理矢量数据实时可视化的高效索引技术

Efficient Indexing Technology for Real‐Time Visualization of Large‐Scale Geographic Vector Data

  • 摘要: 对大规模地理矢量数据进行可视化是当前地理信息科学领域研究的热点与难点,在当前研究中,显示导向型计算方法具有对数据规模不敏感的特性,可支持大规模地理矢量数据实时可视化。但随着数据规模增长,该方法在数据预处理时索引构建时间和生成索引尺寸急剧增长,这大幅影响方法的实用性。基于此,提出一种基于显示导向型计算的高效索引技术:在预处理阶段, 基于四叉树递归划分全球地理空间范围快速构建瓦片四叉(tile-quad, TQ)树索引,在TQ树中通过编码实现结点和瓦片/像素空间范围的对齐;在可视化阶段, 以屏幕显示的像素点为计算单元,判断TQ树中像素对应结点是否存在,可快速计算像素值来生成最终的显示效果。实验结果表明,该技术在数据预处理时的索引构建时间短,索引尺寸小,同时可视化效率远优于现有显示导向可视化方法,可更快地支持亿级矢量要素的实时可视化。

     

    Abstract:
    Objectives Real-time visualization of large-scale geographic vector data is a hot and difficult topic in the field of geographic information science. In the current research, display-driven computing method (DisDC) is insensitive to data scale and can support real-time visualization of large-scale geographic vector data. However, with the growth of data scale, the index construction time and the index size of the DisDC visualization method greatly increase in data preprocessing, which greatly affects the practicability of the method.
    Methods To fill the gap, a fast indexing technique based on DisDC is proposed. Rapid construction of tile-quadtree index (TQ-tree) based on quadtree recursive division of global geographic range in the pre-processing stage, in TQ-tree, the alignment of nodes and tiles/pixels are realized by encoding. In the visualization stage, according to the process of DisDC, the pixel is taken as the calculation unit to determine whether the corresponding node of the pixel in TQ-tree exists, and the pixel value can be quickly calculated to generate the final display effect.
    Results Experimental results show that the proposed technique has shorter index construction time and smaller index size, and the visualization efficiency outperforms the existing DisDC visualization methods.
    Conclusions The Method can support real-time visualization of multi-billion vector elements more quickly.

     

/

返回文章
返回