Journal of Atmospheric and Environmental Optics, Volume. 20, Issue 2, 211(2025)

Remote sensing image classification algorithm combining quadtree structure and scale estimation

FAN Qiang1, WEI Yu1、*, and LUAN Xiaoze2
Author Affiliations
  • 1School of Geomatics, Liaoning Technical University, Fuxin 123000, China
  • 2Shenyang Municipal Engineering Design and Research Institute Co., Ltd, Shenyang 110000, China
  • show less
    Figures & Tables(15)
    Schematic diagram of image segmentation process of fractal network evolution algorithm. (a) Merge step 1; (b) merge step 2
    Schematic diagram of quadtree segmentation algorithm. (a) Process diagram for segmentation; (b) schematic diagram of segmentation
    Flow chart of remote sensing image classification algorithm combining quadtree structure and scale estimation
    Image of the first group of experiments. (a) Building experimental area; (b) wasteland experimental area; (c) road experimental area; (d) water experimental area
    Image change rate-local variance map of experimental area. (a) Building experimental area; (b) wasteland experimental area; (c) road experimental area; (d) water experimental area
    Local variance histogram. (a) Building experimental area; (b) wasteland experimental area; (c) road experimental area; (d) water experimental area
    Variation diagram of classification accuracy in experimental area. (a) Building experimental area; (b) wasteland experimental area
    Second group of experimental image
    Image segmentation result。(a) Overall division result; (b) refined division result 1; (c) refined division result 2
    Schematic diagram of area division by quadtree algorithm. (a) The first partition result; (b) the second division result; (c) the third division result; (d) final division result
    Remote sensing image segmentation result combining quadtree structure and scale estimatio. (a) Quadtree algorithm partitioning result; (b) image segmentation result
    Region segmentation (a) and direct segmentation (b) of classified images
    • Table 1. Optimal segmentation scale of different experimental areas

      View table
      View in Article

      Table 1. Optimal segmentation scale of different experimental areas

      方法最优分割尺度
      建筑实验区荒地实验区道路实验区水体实验区
      ESP方法37313846
      谱空间统计法49183751
    • Table 2. Comparison of classification accuracy between divided region segmentation and direct segmentation

      View table
      View in Article

      Table 2. Comparison of classification accuracy between divided region segmentation and direct segmentation

      实验土地覆盖类型用户精度/%Kappa系数
      建筑物植被水体耕地总体精度
      划分区域分割分类94.2384.6196.8797.9596.800.915
      直接分割分类9482.6196.2996.595.600.879
    • Table 3. Comparison table of classification time between area division and direct division

      View table
      View in Article

      Table 3. Comparison table of classification time between area division and direct division

      分割方法分类时间/s总用时/s
      四叉树分割分割尺度估计FNEA算法
      直接分割0187.5113.70201.21
      本文方法0.67242.2568.36311.28
    Tools

    Get Citation

    Copy Citation Text

    Qiang FAN, Yu WEI, Xiaoze LUAN. Remote sensing image classification algorithm combining quadtree structure and scale estimation[J]. Journal of Atmospheric and Environmental Optics, 2025, 20(2): 211

    Download Citation

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

    Category:

    Received: Mar. 15, 2022

    Accepted: --

    Published Online: May. 30, 2025

    The Author Email: Yu WEI (1404255347@qq.com)

    DOI:10.3969/j.issn.1673-6141.2025.02.009

    Topics