一种水下桩基三维点云的分割和滤波方法  

A segmentation and filtering method for underwater pile foundation 3D point clouds

作　　者：谢伟华 胡皓 卢文龙 劳文欣 尹铁男 XIE Weihua;HU Hao;LU Wenlong;LAO Wenxin;YIN Tienan(Clean Energy Branch of Huaneng International Power Jiangsu Energy Development Co.,Ltd,Nanjing 210015;China Huaneng Group Clean Energy Technology Research Institute Co.,Ltd,Beijing 102209)

机构地区：[1]华能国际电力江苏能源开发有限公司清洁能源分公司,江苏南京210015 [2]中国华能集团清洁能源技术研究院有限公司,北京102209

出　　处：《海洋测绘》2025年第1期53-56,61,共5页Hydrographic Surveying and Charting

基　　金：国家重点研发计划(2018YFB1501305)。

摘　　要：海上风电机组水下桩基长期承受各种荷载和高水流作用,需要有效监测桩基水下部分变形。针对这一需求,给出一种基于三维成像声纳数据的水下桩基三维点云的分割和滤波方法。首先,采用CSF算法、SMRF及欧氏聚类算法进行点云分割,有效提取桩基点云。然后对点云进行格网降采样处理,并使用RANSAC算法建立桩基点云几何模型,从而获取几何参数;最后,采用移动立方体算法重建三维模型以检测桩基异常。实验结果显示,本文方法能从海量点云中有效提取桩基模型,实现桩基三维模型的几何参数计算、重建和评估,为水下桩基检测提供了一种有效的解决方案。The underwater pile foundations of offshore wind turbines endure various loads and high water flow over the long term,necessitating effective monitoring of the underwater deformation of the piles.To address this need,this paper proposes a segmentation and filter method for underwater pile foundations based on three-dimensional imaging sonar data.First,the CSF algorithm,SMRF,and Euclidean clustering algorithm are used for point cloud segmentation to effectively extract pile foundation point clouds.Then,the point clouds are simplified through grid down-sampling,and the pile foundation point clouds are extracted based on the RANSAC algorithm.A geometric model is established to obtain geometric parameters.Finally,the marching cubes algorithm is used to reconstruct the 3D model to detect pile foundation anomalies.The experimental results show that the parameter of geometric shape was calculated precisely,and surface defects and damage states of the piles were observed in the reconstructed 3D models.This method demonstrates good applicability and scalability,providing an effective solution for underwater pile foundation detection.

关 键 词：水下桩基检测 三维成像声纳 点云滤波 点云三维建模 RANSAC算法 移动立方体算法 

分 类 号：P229[天文地球—大地测量学与测量工程]