• Infrared and Laser Engineering
  • Vol. 50, Issue 12, 20210233 (2021)
Ning Li1, Junmin Wang1, Wenjie Si2, and Zexun Geng1、3
Author Affiliations
  • 1School of Information Engineering, Pingdingshan University, Pingdingshan 467000, China
  • 2School of Electrical & Control Engineering, Henan University of Urban Construction, Pingdingshan 467000, China
  • 3Institute of Surveying and Mapping, Information Engineering University, Zhengzhou 450001, China
  • show less
    DOI: 10.3788/IRLA20210233 Cite this Article
    Ning Li, Junmin Wang, Wenjie Si, Zexun Geng. Multi-view SAR target classification method based on principle of maximum entropy[J]. Infrared and Laser Engineering, 2021, 50(12): 20210233 Copy Citation Text show less
    Flowchart of multi-view SAR target classification based on maximum entropy
    Fig. 1. Flowchart of multi-view SAR target classification based on maximum entropy
    Illustration of the targets to be classified
    Fig. 2. Illustration of the targets to be classified
    Classification results under SOC
    Fig. 3. Classification results under SOC
    Comparison of average classification rates under depression angle variance
    Fig. 4. Comparison of average classification rates under depression angle variance
    Average classification rates under noise corruption
    Fig. 5. Average classification rates under noise corruption
    TypeTraining setTest set
    BMP2231193
    BTR70231194
    T72230194
    T62297271
    BRDM2296272
    BTR60254193
    ZSU23/4297272
    D7297272
    ZIL131297272
    2S1297272
    Table 1. Setup of samples under SOC
    MethodAverage classification accuracy
    Proposed99.36%
    Multi-view 198.74%
    Multi-view 299.13%
    Multi-view 399.21%
    CNN99.08%
    Table 2. Comparison of average classificationrates under SOC
    BMP2BDRM2BTR70T72
    Training set231 (Sn_9563)296231230 (Sn_132)
    Test set424 (Sn_812)
    426 (Sn_9566)00571 (Sn_A04)
    427 (Sn_C21)571 (Sn_A05)
    571 (Sn_A07)
    565 (Sn_A10)
    Table 3. Setup of samples for configuration variance
    MethodAverage classification accuracy
    Proposed98.86%
    Multi-view 196.78%
    Multi-view 297.92%
    Multi-view 398.17%
    CNN96.02%
    Table 4. Comparison of average classification rates under configuration variance
    Depression angle/(°)2S1BDRM2ZSU23/4
    Training set17297296297
    Test set30286285286
    45301301301
    Table 5. Setup of samples under depression angle variance
    Ning Li, Junmin Wang, Wenjie Si, Zexun Geng. Multi-view SAR target classification method based on principle of maximum entropy[J]. Infrared and Laser Engineering, 2021, 50(12): 20210233
    Download Citation