Laser & Optoelectronics Progress, Volume. 59, Issue 4, 0428005(2022)

Urban Tree Extraction Method Based on LiDAR Data and Orthophoto

Fan Xie1, Fengbao Yang1、*, and hong Wei2
Author Affiliations
  • 1School of Information and Communication Engineering, North University of China, Taiyuan , Shanxi 030051, China
  • 2School of Systems Engineering, University of Reading, ReadingRG6 6AU, UK
  • show less
    Figures & Tables(7)
    Flowchart of the proposed method
    Part of the roof has a high NDVI
    Vegetation has a low NDVI in shaded area
    Tree extraction results. (a1)‒(f1) NIRRG images; (a2)‒(f2) provided tree real areas; (a3)‒(f3) tree areas extracted by proposed method; (a4) ‒ (f4) evaluation results
    Typical classification error images
    • Table 1. Evaluation index results

      View table

      Table 1. Evaluation index results

      NumberPtPbUtUbOAF1
      Fig. 4(a)0.93410.92910.90320.95220.93120.9184
      Fig. 4(b)0.95780.92470.84340.98100.93450.8970
      Fig. 4(c)0.92060.94310.85500.97020.93710.8866
      Fig. 4(d)0.90930.93590.86270.95880.92770.8854
      Fig. 4(e)0.92090.92500.88100.95100.92350.9005
      Fig. 4(f)0.90900.98040.92140.97710.96600.9152
    • Table 2. Comparison of evaluation indexes of different methods

      View table

      Table 2. Comparison of evaluation indexes of different methods

      MethodOAF1MethodOAF1
      Xiangda180.8980.908Chisheng230.8780.811
      HSN190.8880.883Markus240.8920.823
      HRNet200.9010.889Nan250.8630.846
      C3Net210.9130.903Qiong260.8910.850
      Marmanis220.9030.896Proposed method0.9370.901
    Tools

    Get Citation

    Copy Citation Text

    Fan Xie, Fengbao Yang, hong Wei. Urban Tree Extraction Method Based on LiDAR Data and Orthophoto[J]. Laser & Optoelectronics Progress, 2022, 59(4): 0428005

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category: Remote Sensing and Sensors

    Received: Apr. 8, 2021

    Accepted: May. 18, 2021

    Published Online: Jan. 25, 2022

    The Author Email: Fengbao Yang (yfengb@163.com)

    DOI:10.3788/LOP202259.0428005

    Topics