测绘通报 ›› 2021, Vol. 0 ›› Issue (9): 120-123.doi: 10.13474/j.cnki.11-2246.2021.0287

• 技术交流 • 上一篇    下一篇

基于RANSAC算法的隧道点云横断面提取

丁鸽, 燕立爽, 彭健, 朱传广, 黄雪亭   

  1. 济南市勘察测绘研究院, 山东 济南 250013
  • 收稿日期:2021-03-11 修回日期:2021-07-22 出版日期:2021-09-25 发布日期:2021-10-11
  • 作者简介:丁鸽(1989-),男,硕士,工程师,主要从事工程测量、三维激光扫描方面的研究。E-mail:645474779@qq.com

Cross section extraction of tunnel point cloud based on ransac algorithm

DING Ge, YAN Lishuang, PENG Jian, ZHU Chuanguang, HUANG Xueting   

  1. Jinan Institute of Surveying and Mapping Survey, Jinan 250013, China
  • Received:2021-03-11 Revised:2021-07-22 Online:2021-09-25 Published:2021-10-11

摘要: 针对最小二乘法对所有点(包括"局外点")拟合难以得到最佳线性参数的问题,本文基于RANSAC算法对观测值随机抽样进行参数估算,判断符合模型的"内部点"为一致集,并通过迭代得到足够的一致集,最后设计试验验证该算法的适用性;对隧道点云采用基于中轴线方法进行边界提取,以及三维激光扫描用于生产实践提供参考意义。

关键词: 三维激光扫描, 最小二乘法, 随机采样一致性, 横断面点云, 特征提取

Abstract: Because it is difficult to get the best linear parameters by least square method by fitting all points, including "outliers", this paper presents a method based on RANSAC algorithm to estimate the parameters by random sampling the observed values, the "inner point" of the model is a consistent set until a sufficient consistent set is obtained under a certain number of iterations. The boundary of tunnel point cloud is extracted based on the method of central axis, which provides a reference for the application of 3D laser scanning in production.

Key words: 3D laser scanning, least squares, random sampling consistency, cross-sectional point cloud, feature extraction

中图分类号: