东北大学学报(自然科学版) ›› 2012, Vol. 33 ›› Issue (2): 271-275.DOI: -

• 论著 • 上一篇    下一篇

基于RANSAC模型的机载LiDAR数据中建筑轮廓提取算法

王植;李慧盈;吴立新;贺正雄;   

  1. 东北大学资源与土木工程学院;吉林大学计算机科学与技术学院;北京师范大学减灾与应急管理研究院;
  • 收稿日期:2013-06-19 修回日期:2013-06-19 发布日期:2013-01-17
  • 通讯作者: -
  • 作者简介:-
  • 基金资助:
    国家重点基础研究发展计划项目(2011CB707102);;

Building outline extraction from airborne LiDAR data based on RANSAC model

Wang, Zhi (1); Li, Hui-Ying (2); Wu, Li-Xin (1); He, Zheng-Xiong (1)   

  1. (1) School of Resources and Civil Engineering, Northeastern University, Shenyang 110819, China; (2) College of Computer Science and Technology, Jilin University, Changchun 130012, China; (3) Academy of Disaster Reduction and Emergency Management, Beijing Normal University, Beijing 100875, China
  • Received:2013-06-19 Revised:2013-06-19 Published:2013-01-17
  • Contact: Wang, Z.
  • About author:-
  • Supported by:
    -

摘要: 使用正交多项式分带滤波方法对机载LiDAR点云数据进行滤波处理,通过迭代不断剔除非地面高点数据,最终得到由贴近地面的数据拟合而成的正交多项式.通过设定高程阈值将数据分成地面部分与非地面部分.提出了一种基于随机抽样一致性(RANSAC)算法模型的建筑物面片识别和轮廓提取算法,实现在包含噪声的点云数据中快速准确地识别和提取建筑物轮廓.在实验中对长春市的机载LiDAR数据进行了滤波、建筑屋顶面及其轮廓的提取,验证了本文算法的较高效率和精度.

关键词: 正交多项式滤波, 建筑轮廓, 点云数据, 特征提取, 随机抽样一致性

Abstract: The orthogonal polynomial filtering algorithm was applied to filtering the point cloud data of airborne LiDAR. The non-ground points were excluded continually by iteration and then the orthogonal polynomials were obtained by fitting the points which were close to the ground. The point cloud data were classified into ground and non-ground parts according to the height thresholds. A RANSAC-based building feature extraction algorithm was proposed, which can recognize and extract the information of roof planes and their building outlines fast and accurately from the non-ground point cloud data with noise. The airborne LiDAR data of Changchun city were used to experimentally verify the effectiveness and accuracy of the proposed algorithms.

中图分类号: