Journals Articles Conferences News About CLP
Submit a paper Email Alert
Search
Search
Articles
Journals
News
Advanced Search
Top Searches
2D MaterialsTransformation opticsQuantum PhotonicsSmall lasersSemiconductor UV PhotonicsSupersymmetric microring laser arrays

Acta Optica Sinica, Volume. 44, Issue 20, 2015002(2024)

Algorithm for Eliminating Mismatched Feature Points in Heterogeneous Images Pairs Under Spatial Constraints

Ying Shen, Ye Lin, Haitao Chen, Jing Wu, and Feng Huang*
Author Affiliations
  • School of Mechanical Engineering and Automation, Fuzhou University, Fuzhou 350108, Fujian , China
    • show less
    AbstractGet PDF(in Chinese)
    Figures&Tables (16)
    Equations (0)
    References (31)
    Tools
    Paper Information
    Topics
    Figures & Tables(16)
    SC-PRISAC algorithm flow diagram

    Fig. 1. SC-PRISAC algorithm flow diagram

    Download full size
    Dual band calibration target. (a) Visible light image; (b) infrared image

    Fig. 2. Dual band calibration target. (a) Visible light image; (b) infrared image

    Download full size
    Comparison of images collected from circular hollow target areas. (a) Visible light; (b) infrared

    Fig. 3. Comparison of images collected from circular hollow target areas. (a) Visible light; (b) infrared

    Download full size
    Bilateral filter pyramid sampling process

    Fig. 4. Bilateral filter pyramid sampling process

    Download full size
    Examples of visible light scenes under different illuminances. (a) Low illumination; (b) normal illuminance

    Fig. 5. Examples of visible light scenes under different illuminances. (a) Low illumination; (b) normal illuminance

    Download full size
    Examples of visible light scenes with different depths of field. (a) Close-up view; (b) distant view

    Fig. 6. Examples of visible light scenes with different depths of field. (a) Close-up view; (b) distant view

    Download full size
    Comparison of MRE between our method and spot detection method

    Fig. 7. Comparison of MRE between our method and spot detection method

    Download full size
    Calibration images. (a) Group 2 image; (b) group 9 image

    Fig. 8. Calibration images. (a) Group 2 image; (b) group 9 image

    Download full size
    Polar accuracy results of infrared camera calibration. (a)(b) Spot detection method; (c)(d) our method

    Fig. 9. Polar accuracy results of infrared camera calibration. (a)(b) Spot detection method; (c)(d) our method

    Download full size
    Polar accuracy results of visible light camera calibration. (a)(b) Spot detection method; (c)(d) our method

    Fig. 10. Polar accuracy results of visible light camera calibration. (a)(b) Spot detection method; (c)(d) our method

    Download full size
    Mismatch removal effect picture of SC-PRISAC. (a) Original matching effect; (b) matching effect after spatial constraints; (c) final effect

    Fig. 11. Mismatch removal effect picture of SC-PRISAC. (a) Original matching effect; (b) matching effect after spatial constraints; (c) final effect

    Download full size
    Comparison of experimental results for removing mismatched feature points

    Fig. 12. Comparison of experimental results for removing mismatched feature points

    Download full size
    Comparison of algorithm accuracy under different outer point ratios

    Fig. 13. Comparison of algorithm accuracy under different outer point ratios

    Download full size

    • Table 1. Comparison of average calibration error between our method and spot detection method

      View table

      Table 1. Comparison of average calibration error between our method and spot detection method

      ParameterSpot detection method （OpenCV）Our method
      InfraredVisible lightInfraredVisible light
      Monocular calibration /pixel0.2460.2810.1710.234
      Stereo calibration /pixel1.0280.430

    • Table 2. Display of matching quantity of feature points

      View table

      Table 2. Display of matching quantity of feature points

      Processing stageNumber of interior pointsNumber of external points
      Original matching3054
      Matching after spatial constraints3027
      Final matching2917

    • Table 3. Comparison of evaluation indicators for mismatched feature point removal experiment

      View table

      Table 3. Comparison of evaluation indicators for mismatched feature point removal experiment

      MethodNumber of inliersEstimation error of homography matrix /pixelTime /ms
      RANSAC31.636±15.8939.106±4.17514.752±13.537
      DEGENSAC39.727±21.6598.050±3.8135.297±4.232
      BANSAC35.683±16.6598.605±4.35725.912±10.152
      MAGSAC++15.455±10.59513.924±14.72425.544±10.366
      GC-RANSAC13.091±8.66011.563±8.16523.037±22.836
      ∇-RANSAC（NNs）40.651±19.1637.974±3.26419.613±5.783
      SC-PRISAC（ours）39.182±21.9797.857±3.3591.919±0.629
    Tools

    Get Citation

    Copy Citation Text

    Ying Shen, Ye Lin, Haitao Chen, Jing Wu, Feng Huang. Algorithm for Eliminating Mismatched Feature Points in Heterogeneous Images Pairs Under Spatial Constraints[J]. Acta Optica Sinica, 2024, 44(20): 2015002

    Download Citation

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

    Category: Machine Vision

    Received: Apr. 24, 2024

    Accepted: May. 28, 2024

    Published Online: Oct. 12, 2024

    The Author Email: Huang Feng (huangf@fzu.edu.cn)

    DOI:10.3788/AOS240908

    CSTR:32393.14.AOS240908

    Topics

    laser devices and laser physics
    Lasers and Laser Optics
    Laser physics
    laser manufacturing
    Instrumentation, Measurement and Metrology