基于四叉树的多波束水深数据快速检索方法  

A Fast Retrieval Method for Multibeam Sounding Data Processing Based on Quadtree

作　　者：孟浩 徐卫明[1] 刘天阳 MENG Hao;XU Weiming;LIU Tianyang(Department of Military Oceanography and Hydrography,Dalian Naval Academy,Dalian 116018,China)

机构地区：[1]海军大连舰艇学院军事海洋与测绘系,辽宁大连116018

出　　处：《海洋测绘》2020年第6期25-28,33,共5页Hydrographic Surveying and Charting

基　　金：国家自然科学基金(61071006)。

摘　　要：针对水下地形测量仿真器中的多波束测深结果仿真环节中,对海量已知水深点云数据进行处理时存在的检索效率低、内存占用大的问题,设计了一种基于四叉树的数据处理方案。以四叉树数据结构并采用序列化方式存储索引文件,提高点云数据的检索效率;通过内存映射的方法读取海量点云数据,减少内存占用。数据实验表明:相对于常见的遍历检索,在检索点云数不超过总点云数的约四分之三时,四叉树水深点云检索效率提高了1倍以上;在检索点云数越少,总点数云越多时,四叉树水深点云检索的效率最多可提升30倍以上。基于四叉树的数据处理算法可有效地提高点云数据的检索效率,适合于多波束测深结果仿真过程中的海量点云数据的处理。Aiming at the problem of low retrieval efficiency and large memory consumption when processing the mass known bathymetry point cloud data during the simulation of multibeam echou sounding system results in underwater terrain surveying simulator,a data processing scheme based on quadtree is designed.Index files are stored in a quadtree data structure and serialized to improve the retrieval efficiency of cloud data.Massive point cloud data are read through memory mapping to reduce memory footprint.Experiment shows that,compared with common traversal retrieval,the quadtree retrieval efficiency of point cloud data is more than one time on condition that the number of retrieval point cloud data does not exceed about three quarters of the total number of point cloud data.Furthermore,the quadtree retrieval efficiency of point cloud data can be improved by more than 30 times at most,while having a larger number of total point cloud data and fewer retrieval point cloud data.The data processing algorithm based on quadtree can effectively improve the retrieval efficiency of point cloud data retrieval and is suitable for the processing of mass known multibeam point cloud data.

关 键 词：水深测量 水下地形测量仿真 多波束测深系统 数据检索 四叉树 

分 类 号：P229.3[天文地球—大地测量学与测量工程]