基于中轴线拟合的隧道点云去噪研究
|
摘要
目的通过拟合中轴线对隧道内部噪声点滤除,并在去噪时保持隧道主体数据的完整性,提出一种通过拟合隧道中轴线对隧道点云进行噪声点滤除的方法.方法采用统计滤波滤除部分远离主体点云的大尺度噪声点,再基于高斯曲率的方法将隧道点云数据分为管片点云和轨道面点云两类;利用投影法提取隧道中轴线,通过判断点到中轴线之间的距离与设定的阈值的大小关系,滤除隧道内部及靠近管片壁的噪声点,多次迭代去噪,直到点云数据量趋于稳定;利用直通滤波和统计滤波滤除轨道面点云噪声,最后将去噪后的管片点云和轨道面点云进行组合得到去噪后的盾构隧道点云.结果利用拟合得到的隧道中轴线对分类后的管片点云进行五次迭代去噪后,管片点云数量趋于稳定,噪声点基本滤除;对分类后的轨道面点云进行了有效滤除并且保持了原有特征;将去噪后的两部分点云组合,得到了完整的去噪后的隧道点云.结论笔者所采用的方法简单有效,能有效地滤除隧道点云噪声且保证主体数据完整性.
To filter the inner noise points in the tunnel by the method of fitting the central axis,and keep the integrity of the tunnel main data while denoising,a method to filter the noise points of the tunnel point cloud by fitting the central axis of the tunnel was proposed.Firstly,statistical filtering was applied to filter some large-scale noise points farther from the main point cloud.Then the Gauss curvature method was adopted to divided point cloud data of tunnel into two categories:the point cloud of segment and the point cloud of orbital plane.The central axis of the tunnel was extracted by the method of projection.By judging the relationship between the distance from the point to the central axis and the set threshold,noise points inner tunnel and near the segment was to be filtered out,and with multiple iterations for denoising until the amount of the point cloud data tends to be stable.After that,the method of straight-through filtering and statistical filtering were used to filter the point cloud noise on the orbital surface.Finally,the denoised segment point cloud and the orbital surface point cloud were combined to obtain the shield tunnel point cloud after denoising.The results show that after five iterations denoising for the classified segmented point clouds by using the fitting tunnel central axis,the number of segmented point clouds tends to be stable and the noise points are eliminated;the classified orbital plane point clouds are effectively filtered and retained their original characteristics;combining the two parts of the denoised point clouds,the complete denoised tunnel point clouds was obtained.This method,simple and effective,can effectively filter the tunnel point cloud noise and ensure the integrity of the main data.
To filter the inner noise points in the tunnel by the method of fitting the central axis,and keep the integrity of the tunnel main data while denoising,a method to filter the noise points of the tunnel point cloud by fitting the central axis of the tunnel was proposed.Firstly,statistical filtering was applied to filter some large-scale noise points farther from the main point cloud.Then the Gauss curvature method was adopted to divided point cloud data of tunnel into two categories:the point cloud of segment and the point cloud of orbital plane.The central axis of the tunnel was extracted by the method of projection.By judging the relationship between the distance from the point to the central axis and the set threshold,noise points inner tunnel and near the segment was to be filtered out,and with multiple iterations for denoising until the amount of the point cloud data tends to be stable.After that,the method of straight-through filtering and statistical filtering were used to filter the point cloud noise on the orbital surface.Finally,the denoised segment point cloud and the orbital surface point cloud were combined to obtain the shield tunnel point cloud after denoising.The results show that after five iterations denoising for the classified segmented point clouds by using the fitting tunnel central axis,the number of segmented point clouds tends to be stable and the noise points are eliminated;the classified orbital plane point clouds are effectively filtered and retained their original characteristics;combining the two parts of the denoised point clouds,the complete denoised tunnel point clouds was obtained.This method,simple and effective,can effectively filter the tunnel point cloud noise and ensure the integrity of the main data.
引文
[1] 肖国新,于辉,周晓辉.一种自适应双边滤波点云去噪算法[J].信息技术,2017(10):118-120.(XIAO Guoxin,YU Hui,ZHOU Xiaohui.An adaptive bilateral filtering point cloud denoising algorithm[J].Information technology,2017(10):118-120.)
[2] 荆方,刘增力.基于改进的双边滤波与非下采样剪切波变换的图像去噪[J].通信技术,2019,52(1):80-85.(JING Fang,LIU Zengli.Image denoising based on improved bilateral filtering and non-subsampled shearlet transform[J].Communications technology,2019,52(1):80-85.)
[3] CASTILLO E,LIANG J,ZHAO H.Point cloud segmentation and denoising via constrained nonlinear least squares normal estimates [M].Germany:Springer Berlin Heidelberg,2013:283-299.
[4] 周浩宇,辛国江,谭光华,等.基于脉冲耦合神经网络的点云曲面去噪[J].电子学报,2012,40(11):2221-2225.(ZHOU Haoyu,XIN Guojiang,TAN Guanghua,et al.PCNN-based point set surface denoising[J].Acta electronica sinica,2012,40(11):2221-2225.)
[5] WANG X,LI Z,MAI Y,et al.Robust denoising of unorganized point clouds[C]//Proceedings of the 2011 international conference on intelligent computing and integrated systems.Piscatawway:IEEE,2013:1-3.
[6] 曹红新.机载LIDAR数据滤波方法研究[D].成都:西南交通大学,2011.(CAO Hongxin.The method study of LIDAR filtering[D].Chengdu:Southwest Jiaotong University,2011.)
[7] 吴丛丛,卢小平,李国立,等.基于TIN的LiDAR数据滤波算法研究[J].测绘通报,2013(3):32-35.(WU Congcong,LU Xiaoping,LI Guoli,et al.Research on filtering algorithm for LiDAR data based on TIN[J].Bulletin of surveying and mapping,2013(3):32-35.)
[8] 张靖,张晓君,江万寿,等.一种改进的线性预测滤波算法[J].国土资源遥感,2011(1):52-56.(ZHANG Jing,ZHANG Xiaojun,JIANG Wanshou,et al.Progressive linear prediction fitting for extracting DTM from airborne LiDAR Data [J].Remote sensing for land & resources,2011(1):52-56.)
[9] 高志国.海量点云数据滤波处理方法研究[J].测绘工程,2013,22(1):35-38.(GAO Zhiguo.Research on the massive point cloud data filtering process method[J].Engineering of surveying and mapping,2013,22(1):35-38.)
[10] 余鹏磊,于海洋,谢秋平,等.基于RANSAC算法的地铁隧道点云数据滤波[J].测绘工程,2014,23(8):28-31.(YU Penglei,YU Haiyang,XIE Qiuping,et al.Filtering for subway tunnel point cloud data based on RANSAC algorithm[J].Engineering of surveying and mapping,2014,23(8):28-31.)
[11] 邓辉,蓝秋萍,廖威,等.基于法向量偏差的隧道点云去噪算法[J].测绘工程,2018,27(1):59-63.(DENG Hui,LAN Qiuping,LIAO Wei,et al.Tunnel point clouds denoising algorithm based on normal deviation[J].Engineering of surveying and mapping,2018,27(1):59-63.)
[12] 琚俏俏,程效军,徐工.基于椭圆拟合的隧道点云去噪方法[J].工程勘察,2014,42(9):69-72.(JU Qiaoqiao,CHENG Xiaojun,XU Gong.Denoising of tunnel point cloud data based on ellipse fitting[J].Geotechnical investigation & surveying,2014,42(9):69-72.)
[13] 程效军,贾东峰,刘艳萍,等.基于中轴线的隧道点云去噪[J].同济大学学报(自然科学版),2015,48(8):1239-1248.(CHENG Xiaojun,JIA Dongfeng,LIU Yanping,et al.Tunnel point cloud denoising algorithm based on centerline[J].Journal of tongji university(natural science),2015,48(8):1239-1248.)
[14] 徐光华.基于二次曲线拟合的隧道激光点云滤波方法及应用[J].测绘通报,2015(5):42-45,108.(XU Guanghua.Laser point cloud filtering and application in tunnel deformation monitoring based on quadratic curves fitting[J].Bulletin of surveying and mapping,2015(5):42-45,108.)
[15] 陈璋雯,达飞鹏.基于模糊熵迭代的三维点云精简算法[J].光学学报,2013,33(8):161-167.(CHEN Zhangwen,DA Feipeng.3D point cloud simplification algorithm based on fuzzy entropy iteration[J].Acta optica sinica,2013,33(8):161-167.)
[16] 王银腾.基于包围盒和最大信息相似性误差测度的点云精简[D].秦皇岛:燕山大学,2017.(WANG Yinteng.Point cloud simplification based on bounding box and maximum information similarity error measure[D].Qinhuangdao:Yanshan University,2017.)
[17] JALBA A C,ROERDINK J B T M.Efficient surface reconstruction from noisy data using regularized membrane potentials[J].IEEE transactions on image processing,2009,18(5):1119-1134.
[18] JIANG J Y,ZHANG Z X,MING Y.The airborne Lidar points cloud filtering of complex city environment [J].Journal of Wuhan university,2007(5):402-405.
[2] 荆方,刘增力.基于改进的双边滤波与非下采样剪切波变换的图像去噪[J].通信技术,2019,52(1):80-85.(JING Fang,LIU Zengli.Image denoising based on improved bilateral filtering and non-subsampled shearlet transform[J].Communications technology,2019,52(1):80-85.)
[3] CASTILLO E,LIANG J,ZHAO H.Point cloud segmentation and denoising via constrained nonlinear least squares normal estimates [M].Germany:Springer Berlin Heidelberg,2013:283-299.
[4] 周浩宇,辛国江,谭光华,等.基于脉冲耦合神经网络的点云曲面去噪[J].电子学报,2012,40(11):2221-2225.(ZHOU Haoyu,XIN Guojiang,TAN Guanghua,et al.PCNN-based point set surface denoising[J].Acta electronica sinica,2012,40(11):2221-2225.)
[5] WANG X,LI Z,MAI Y,et al.Robust denoising of unorganized point clouds[C]//Proceedings of the 2011 international conference on intelligent computing and integrated systems.Piscatawway:IEEE,2013:1-3.
[6] 曹红新.机载LIDAR数据滤波方法研究[D].成都:西南交通大学,2011.(CAO Hongxin.The method study of LIDAR filtering[D].Chengdu:Southwest Jiaotong University,2011.)
[7] 吴丛丛,卢小平,李国立,等.基于TIN的LiDAR数据滤波算法研究[J].测绘通报,2013(3):32-35.(WU Congcong,LU Xiaoping,LI Guoli,et al.Research on filtering algorithm for LiDAR data based on TIN[J].Bulletin of surveying and mapping,2013(3):32-35.)
[8] 张靖,张晓君,江万寿,等.一种改进的线性预测滤波算法[J].国土资源遥感,2011(1):52-56.(ZHANG Jing,ZHANG Xiaojun,JIANG Wanshou,et al.Progressive linear prediction fitting for extracting DTM from airborne LiDAR Data [J].Remote sensing for land & resources,2011(1):52-56.)
[9] 高志国.海量点云数据滤波处理方法研究[J].测绘工程,2013,22(1):35-38.(GAO Zhiguo.Research on the massive point cloud data filtering process method[J].Engineering of surveying and mapping,2013,22(1):35-38.)
[10] 余鹏磊,于海洋,谢秋平,等.基于RANSAC算法的地铁隧道点云数据滤波[J].测绘工程,2014,23(8):28-31.(YU Penglei,YU Haiyang,XIE Qiuping,et al.Filtering for subway tunnel point cloud data based on RANSAC algorithm[J].Engineering of surveying and mapping,2014,23(8):28-31.)
[11] 邓辉,蓝秋萍,廖威,等.基于法向量偏差的隧道点云去噪算法[J].测绘工程,2018,27(1):59-63.(DENG Hui,LAN Qiuping,LIAO Wei,et al.Tunnel point clouds denoising algorithm based on normal deviation[J].Engineering of surveying and mapping,2018,27(1):59-63.)
[12] 琚俏俏,程效军,徐工.基于椭圆拟合的隧道点云去噪方法[J].工程勘察,2014,42(9):69-72.(JU Qiaoqiao,CHENG Xiaojun,XU Gong.Denoising of tunnel point cloud data based on ellipse fitting[J].Geotechnical investigation & surveying,2014,42(9):69-72.)
[13] 程效军,贾东峰,刘艳萍,等.基于中轴线的隧道点云去噪[J].同济大学学报(自然科学版),2015,48(8):1239-1248.(CHENG Xiaojun,JIA Dongfeng,LIU Yanping,et al.Tunnel point cloud denoising algorithm based on centerline[J].Journal of tongji university(natural science),2015,48(8):1239-1248.)
[14] 徐光华.基于二次曲线拟合的隧道激光点云滤波方法及应用[J].测绘通报,2015(5):42-45,108.(XU Guanghua.Laser point cloud filtering and application in tunnel deformation monitoring based on quadratic curves fitting[J].Bulletin of surveying and mapping,2015(5):42-45,108.)
[15] 陈璋雯,达飞鹏.基于模糊熵迭代的三维点云精简算法[J].光学学报,2013,33(8):161-167.(CHEN Zhangwen,DA Feipeng.3D point cloud simplification algorithm based on fuzzy entropy iteration[J].Acta optica sinica,2013,33(8):161-167.)
[16] 王银腾.基于包围盒和最大信息相似性误差测度的点云精简[D].秦皇岛:燕山大学,2017.(WANG Yinteng.Point cloud simplification based on bounding box and maximum information similarity error measure[D].Qinhuangdao:Yanshan University,2017.)
[17] JALBA A C,ROERDINK J B T M.Efficient surface reconstruction from noisy data using regularized membrane potentials[J].IEEE transactions on image processing,2009,18(5):1119-1134.
[18] JIANG J Y,ZHANG Z X,MING Y.The airborne Lidar points cloud filtering of complex city environment [J].Journal of Wuhan university,2007(5):402-405.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |