| 网格总分并行式Delaunay三角网建模方法 |
| |
| 引用本文： | 韩元利.网格总分并行式Delaunay三角网建模方法[J].测绘学报,2015,44(6):702-708. |
| |
| 作者姓名： | 韩元利 |
| |
| 作者单位： | 中铁第四勘察设计院集团有限公司, 湖北 武汉 430063 |
| |
| 摘 要： | 针对大规模点云数据,提出了Delaunay三角网构建的一种算法,算法通过自适应网格空间分割,实现了海量点云数据的规模均衡网格化逻辑分割;对网格内的顶点按距中距离进行排序,通过各网格由外而内的插入法建立三角网;按先总后分的方式优先保障网格之间三角网的生成,避免了分治-综合建模算法复杂而低效的三角网整合过程;建立了网格的拓扑闭包检测机制,针对各个子网格适时启动独立并行的线程对余下的内部点按传统的拓扑插入算法进行独立建模,从而并行高效、由总到分地实现了海量点集数据的三角网建模工作,显著地提高了空间大数据的三角网建模能力。
|
| 关 键 词： | 三角网建模 总分式构TIN Delaunay三角网 多线程构TIN 大数据 |
| 收稿时间： | 2014-03-13 |
| 修稿时间： | 2014-12-16 |
A General-division Grid Pattern Delaunay-TIN Paral lel Algorithm |
| |
| HAN Yuan Li.A General-division Grid Pattern Delaunay-TIN Paral lel Algorithm[J].Acta Geodaetica et Cartographica Sinica,2015,44(6):702-708. |
| |
| Authors: | HAN Yuan Li |
| |
| Institution: | China Railway Siyuan Survey and Design Group Co., LTD, Wuhan 430063, China |
| |
| Abstract: | This paper achieves out a new Delaunay triangulation algorithm.Fi rstly,the self-adaptation grid space division was proposed to real ize the balanced logical grid division for massive point cloud data. Secondly,from far to near order the sequence of points in each grid by distance to the grid center and find out the nearest point and mark it as the central point.Thi rdly,the TIN was bui lt with by a new general-division Delaunay triangulation algorithm,which uses traditional insertion method to bui ld TIN and add only one point from each grid at one times to form new TIN.When bui lding TIN we use find-insertion method fi rstly and hereafter use topology-insertion method to keep high efficiency.This algorithm has good efficiency because it successful ly avoided the merge process of sub grid triangulation mesh.Final ly,the topological closure detection mechanism was establ ished,and the independent paral lel multithreading was started to model the rest points by topology-insertion algorithm l imit to every grid space,which made the triangulation model ing of the whole space efficient.The method of this paper improved the support capacity of space model ing for massive point cloud data obviously. |
| |
| Keywords: | triangulation model ing general-division pattern TIN construction Delaunay triangulation multithreading construction massive data |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《测绘学报》浏览原始摘要信息 |
| 点击此处可从《测绘学报》下载免费的PDF全文 |
|
