海量三维表面模型快速切片方法研究  

Research on Fastly Slicing of Massive 3D Surface Model

作　　者：方园园 付仲良[1] 李玉舟 张涵 FANG Yuanyuan;FU Zhongliang;LI Yuzhou;ZHANG Han(School of Remote Sensing and Information Engineering,Wuhan University,Wuhan 430079,China;Highway Administration Bureau of Henan Transportation Department,Zhengzhou 450000,China)

机构地区：[1]武汉大学遥感信息工程学院,湖北武汉430079 [2]河南省交通运输厅公路管理局,河南郑州450000

出　　处：《测绘地理信息》2023年第4期7-10,共4页Journal of Geomatics

基　　金：基于大数据分析的公路视频监控与感知成套技术研究(2019G1)。

摘　　要：针对大规模场景中三维数据的快速加载,对原始三维数据进行切片处理,是解决海量数据加载调度的关键技术手段之一。提出一种基于GPU(graphics processing unit)并行的三维表面模型快速切片方法,通过并行判断三角形与包围盒的空间关系,计算三角形与包围盒的交点并重新构建三角网,最后得到切片数据。通过实验验证,该方法能有效提高海量模型的切片效率。For the fast loading of 3D data in large-scale scenes,slice of original 3D data is one of the key technical means to solve massive data loading scheduling.This paper proposes a fast slicing method for 3D surface models based on GPU parallelism.By judging the spatial relationship between the triangle and the bounding box in parallel,calculating the intersection point of the triangle and the bounding box and rebuilding the triangulation network,the slice data is finally obtained.It is verified by experiment that this method can effectively improve the slicing efficiency of massive models.

关 键 词：三角网分割 三维瓦片数据集 GPU并行计算 

分 类 号：P208[天文地球—地图制图学与地理信息工程] P237[天文地球—测绘科学与技术]