基于密度聚类和圆外切线斜率拟合的埋地排水管道声呐点云去噪技术  

The Slope Fitting Method of Sonar Point Cloud on the Inner Wall of Sewer Based on Density Algorithm and Tangent Slope Fitting of the Circle

作　　者：吴俊豪 李琛 王蓉 刘文黎 周星宇 WU Jun-hao;LI Chen;WANG Rong;LIU Wen-li;ZHOU Xing-yu(School of Civil Engineering&Mechanics,Huazhong University of Science&Technology,Wuhan,Hubei 430074,China)

机构地区：[1]华中科技大学土木工程与力学学院,湖北武汉430074

出　　处：《计量学报》2023年第2期231-237,共7页Acta Metrologica Sinica

摘　　要：针对环扫声呐扫描结果受到排水管道内混合水声的影响问题,提出一种声呐扫描点云数据去噪的方法:在环扫声呐扫描得到的三维点云数据基础上,通过密度聚类优化算法,去除点云数据中的噪声,然后采用类圆外切线斜率拟合的方式,识别出管道内壁界限和淤泥淤积线,最终得出包含排水管道内壁界限和淤积线特征的模型。为验证有效性,以武汉市某排水管道为例进行分析,基于现场采集的98万个点云坐标进行数据去噪和特征提取,结果表明:采用密度聚类优化算法进行点云数据初筛后,通过圆外切线斜率拟合算法能有效识别出排水管道内壁界限和淤积线,拟合半径均方误差0.0071 m,相较于单一密度聚类算法拟合精度更高,去噪效果更好。To solve the problem that the scan results of the ring scan sonar are affected by the sound of mixed water in the drainage pipe,a method for denoising the point cloud data of the sonar scan is proposed.On the basis of the three-dimensional point cloud data obtained by the ring scan sonar,the density clustering optimization algorithm(DBSCAN)is used to remove the noise in the point cloud data.Then,the method of fitting the slope of the outer tangent of the circle is used to identify the boundary of the inner wall of the pipeline and the silt deposition line,and finally obtain a model including the boundary of the inner wall of the drainage pipe and the characteristics of the silt line.To verify the effectiveness,a drainage pipeline in Wuhan is taken as an example for analysis,and performs data denoising and feature extraction based on 980000 point cloud coordinates collected on site.The results show that after the initial screening of the point cloud data using the density clustering optimization algorithm,the SFA-COT can effectively identify the inner wall boundary and the siltation line of the drainage pipe.The mean square error of the fitting radius is 0.0071 m,which has higher fitting accuracy and better denoising effect than the single density clustering algorithm.

关 键 词：计量学 声呐扫描 点云数据 去噪 排水管道 密度聚类 类圆拟合 管道特征拟合 

分 类 号：TB95[一般工业技术—计量学] TB973[机械工程—测试计量技术及仪器]