论文部分内容阅读
针对激光雷达三维点云数据量大,当计算机内存有限时进行点云读取与处理存在严重滞后的问题,提出了一种叠加型金字塔索引结构。首先,采用一种基于点云最小外包络的不均匀分块策略,将点云数据划分成若干独立的数据块;待分块完成后,利用提出的叠加型索引结构对每个分块构建金字塔;最后,将生成的金字塔按照指定的文件结构存储,生成索引文件。利用机载实测点云数据开展了验证实验,结果表明:该算法有效地降低了索引文件占据的计算机空间资源,实现了海量三维点云数据在有限内存空间的快速显示。
Aiming at the large data volume of 3D point cloud in LIDAR, there is a serious lag in the point cloud reading and processing when the computer memory is limited, and an overlay type pyramid indexing structure is proposed. First, a non-uniform segmentation strategy based on the minimum envelope of point cloud is used to divide the point cloud data into several independent data blocks. After the block is completed, the proposed overlay index structure is used to construct each block Pyramid; Finally, the generated pyramid according to the specified file structure, generate index file. The experimental results show that this algorithm effectively reduces the computer space resources occupied by the index file and realizes the rapid display of the massive 3D point cloud data in the limited memory space.