Optics and Precision Engineering, Volume. 31, Issue 21, 3156(2023)

Target detection using spectral unmixing

Lei ZHANG1... Kai QIAO1, Yinhua WU2,* and Siyuan LI3 |Show fewer author(s)
Author Affiliations
  • 1Beijing Institute of Tracking and Telecommunication Technology, Beijing00094, China
  • 2Xi'an Technological University, Xi'an71001, China
  • 3Xi'an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi'an710119, China
  • show less
    Figures & Tables(13)
    Flowchart of optimized weighted CEM target detection algorithm using spectral unmixing
    Endmember spectra of simulated hyperspectral image
    Simulated hyperspectral image (Band 70th)
    Endmember labeled image of simulated hyperspetral image
    RGB color map of San Diego airport
    Target groundtruth of San Diego airport
    False color map of Camouflage vehicle
    Target groundtruth of Camouflage vehicle
    Target detection results of different algorithms
    ROC curves of different algorithms
    • Table 1. AUC of different algorithms in experiment 1

      View table
      View in Article

      Table 1. AUC of different algorithms in experiment 1

      算 法模拟-端元1模拟-端元2模拟-端元3模拟-端元4圣地亚哥迷彩车辆平均值
      CEM0.866 40.792 00.809 60.992 00.871 80.951 10.880 5
      基于SAM的加权CEM0.916 30.783 90.835 60.996 20.990 50.998 40.920 2
      基于丰度系数的加权CEM0.921 20.830 20.860 90.996 30.987 00.982 00.929 6
      基于光谱解混合的加权CEM0.921 20.825 10.859 20.996 20.993 70.998 00.932 2
    • Table 2. AUC of different algorithms in experiment2

      View table
      View in Article

      Table 2. AUC of different algorithms in experiment2

      高光谱数据SAM(si'丰度系数(qi'

      SAM+丰度系数((qi'+

      si')/2)

      模拟-端元10.875 50.963 90.959 7
      模拟-端元20.533 80.901 80.859 9
      模拟-端元30.667 30.794 70.824 8
      模拟-端元40.995 70.996 60.996 2
      圣地亚哥0.979 30.992 30.997 2
      迷彩车辆0.997 80.971 00.998 1
      平 均0.841 60.936 70.939 3
    • Table 3. AUC of this algorithm

      View table
      View in Article

      Table 3. AUC of this algorithm

      高光谱数据加权CEMSAM+丰度系数((qi+ si)/2)本文算法
      模拟-端元10.921 20.959 70.966 9
      模拟-端元20.825 10.859 90.877 4
      模拟-端元30.859 20.824 80.873 7
      模拟-端元40.996 20.996 20.996 3
      圣地亚哥0.993 70.997 20.997 5
      迷彩车辆0.998 00.998 10.998 4
      平均0.932 20.939 30.951 7
    Tools

    Get Citation

    Copy Citation Text

    Lei ZHANG, Kai QIAO, Yinhua WU, Siyuan LI. Target detection using spectral unmixing[J]. Optics and Precision Engineering, 2023, 31(21): 3156

    Download Citation

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

    Category:

    Received: Jun. 21, 2023

    Accepted: --

    Published Online: Jan. 5, 2024

    The Author Email: WU Yinhua (yinhuawoo@163.com)

    DOI:10.37188/OPE.20233121.3156

    Topics