首页 | 本学科首页   官方微博 | 高级检索  
     检索      

多波束点云滤波算法初步研究及适用性分析
引用本文:崔晓冬,沈蔚,帅晨甫,惠笑.多波束点云滤波算法初步研究及适用性分析[J].海洋测绘,2021(5):12-16.
作者姓名:崔晓冬  沈蔚  帅晨甫  惠笑
作者单位:上海海洋大学 海洋科学学院,上海 201306 ;上海河口海洋测绘工程技术研究中心,上海 201306;国家海洋局东海信息中心,上海 200137;中船勘察设计研究院有限公司,上海 200063
基金项目:上海市科委重点科研计划(17DZ1204902)
摘    要:为提高多波束数据处理的准确性和简便性,研究了多波束水深点云处理的策略和关键算法,构建了kd-tree(k-dimensional tree)结构加速点云查找,针对大尺度噪点设计了半径滤波和统计滤波算法,针对小尺度噪点提出了双边滤波算法,并对以上算法的适用性进行了探讨分析。实验结果表明,当设置适合的滤波阈值,半径滤波和统计滤波能高效去除大尺度噪点,且保留地形特征,总误差分别为2.25%和2.76%。在大尺度噪点去除的基础上,改进的双边滤波可以实现地形平滑的同时,保持水深点云的有效数据量。研究成果为多波束点云数据的自动化处理提供了解决方案,对多波束测量工作的效果和效率提升做了有益尝试。

关 键 词:滤波算法  半径滤波  统计滤波  双边滤波  多波束点云数据

Preliminary research and application analysis ofmulti-beam point cloud filtering algorithm
CUI Xiaodong,SHEN Wei,SHUAI Chenfu,HUI Xiao.Preliminary research and application analysis ofmulti-beam point cloud filtering algorithm[J].Hydrographic Surveying and Charting,2021(5):12-16.
Authors:CUI Xiaodong  SHEN Wei  SHUAI Chenfu  HUI Xiao
Abstract:In order to improve the accuracy and simplicity of multi-beam data processing,the strategy and key algorithm of multi-beam point cloud processing are designed,and the KD-Tree (K-Dimensional Tree) structure is constructed to accelerate point cloud searching.Radius filtering and statistical filtering algorithms are designed for large scale noise,and bilateral filtering algorithm is designed for small scale noise.At the same time,these algorithms applicability is discussed and analyzed.Experimental results show that when a suitable filtering threshold is set,radius filtering and statistical filtering can efficiently remove large-scale noise and retain topographic features,with a total error of 2.25% and 2.76%,respectively.On the basis of large-scale noise removal,improved bilateral filtering can maintain the data volume of the water depth point while smoothing the terrain.The research results of this paper provide a solution for the automatic processing of multi-beam point cloud data,and make a beneficial attempt to improve the effect and efficiency of multi-beam sonar surveying.
Keywords:
点击此处可从《海洋测绘》浏览原始摘要信息
点击此处可从《海洋测绘》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号