Laser & Optoelectronics Progress, Volume. 62, Issue 12, 1212005(2025)

RANSAC-Based Random Circle Detection and Fitness Evaluation Method for Spot Localization

Xuancai Lu and Jinsong Wang*
Author Affiliations
  • School of Optoelectronic Engineering, Changchun University of Science and Technology, Changchun 130022, Jilin , China
  • show less
    Figures & Tables(14)
    Schematic diagram of the measurement principle of the lensometer
    Influence of pixel displacement on diopter error
    Flowchart of spot localization for proposed algorithm
    Comparison of kernel functions for three interpolation algorithms
    Grey images of Gaussian spots before and after interpolation. (a) Before interpolation; (b) after interpolation
    Comparison of 3D Gaussian surfaces before and after interpolation. (a) Before interpolation; (b) after interpolation
    Binary image of Gaussian spot
    Spot edge extracted by Canny algorithm
    Schematic diagram of radial scanning
    Influence of radial scanning times on localization error and time
    Influence of random circle detection success threshold on localization error and time
    • Table 1. Definitions and preset values of symbols

      View table

      Table 1. Definitions and preset values of symbols

      SymbolDefinitionPreset value
      N'Number of edge points of the spot0
      NscanRadial scanning times100
      Rscan /pixelRadial scanning radius60
      nsNumber of successful random circle detection0
      Tn, sThreshold for successful random circle detection30
      ieNumber of random circle detection executed0
      Ti, eMaximum number of random circle detection5000
      TdistThreshold for the distance from inlier to the random circle2
      gRandom circle fitness0
      TgThreshold for circle fitness0.7
      gmaxCandidate circle fitness0
    • Table 2. Parameter descriptions and preset values of the five methods

      View table

      Table 2. Parameter descriptions and preset values of the five methods

      MethodParameterSymbolPreset value
      RFEThreshold for the distance from inlier to the random circleTdist2
      Maximum number of random circle detectionTi, e5000
      Threshold for successful random circle detectionTn, s10
      Threshold for circle fitnessTg0.7
      Radial scanning radiusRscan /pixel60
      Radial scanning timesNscan100
      RDSDistance threshold for inlierTdist'2
      Number of inlierTin50
      Threshold for random circle sampling timesTfail500
      HTRadius quantization unitΔR /pixel1
      Minimum radiusRmin /pixel30
      Maximum radiusRmax /pixel60
      LSNo predefined parameterNULLNULL
      SWNo predefined parameterNULLNULL
    • Table 3. Average errors and localization time of different methods for spot center localization

      View table

      Table 3. Average errors and localization time of different methods for spot center localization

      MethodAverage localization error /pixelAverage localization time /msAverage diopter error /D
      RFE0.242330.0079
      RDS3.3341560.1100
      HT0.2797430.0092
      LS2.094270.0690
      SW0.757790.0240
    Tools

    Get Citation

    Copy Citation Text

    Xuancai Lu, Jinsong Wang. RANSAC-Based Random Circle Detection and Fitness Evaluation Method for Spot Localization[J]. Laser & Optoelectronics Progress, 2025, 62(12): 1212005

    Download Citation

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

    Category: Instrumentation, Measurement and Metrology

    Received: Dec. 5, 2024

    Accepted: Jan. 6, 2025

    Published Online: Jun. 9, 2025

    The Author Email: Jinsong Wang (soldier_1973@163.com)

    DOI:10.3788/LOP242372

    Topics