Laser & Optoelectronics Progress, Volume. 58, Issue 24, 2415006(2021)

Positioning Algorithm of Homonymous Object Points in Multi-Vision System Based on Weighted Levenberg-Marquardt

Aiguo Zhou, Jiangyang Yu*, Jiakun Pu, Zhenyu Li**, and Ronghui Wei
Author Affiliations
  • School of Mechanical Engineering, Tongji University, Shanghai 201804, China
  • show less
    Figures & Tables(12)
    Visual imaging model
    Ideal multi-vision imaging system
    Real multi-vision imaging system
    Real vision measurement model
    Simplified vision measurement model
    Top view of experiment scene
    Comparison of pixel error
    Comparison of truth error
    Algorithm convergence curve
    • Table 1. Calibration results of internal parameters

      View table

      Table 1. Calibration results of internal parameters

      Cameraf /mmu0 /pixelv0 /pixelk1 /10-5k2 /10-8p1 /10-6p2 /10-6
      C18.047821311.726979.6747.65049.19071.96087.3531
      C28.067591289.349936.0218.37719.46794.75392.6350
      C38.073611257.554941.2539.08455.28633.69314.5361
      C48.054701295.348967.6927.95848.03886.84723.7469
    • Table 2. Calibration results of external parameters

      View table

      Table 2. Calibration results of external parameters

      CameraExternal parameter matrix
      C1-0.6424-0.0262-0.76593868.3 0.6957 0.3993-0.59712798.9 0.3215-0.9164-0.23832704.80001
      C2-0.5712-0.2832 0.7704-3942.6-0.8045 0.0071-0.5939 2876.2 0.1627-0.9590-0.2319 2649.70001
      C3 0.5606-0.0219 0.8278-3623.2-0.7669-0.3909 0.5090-2835.7 0.3124-0.9202-0.2360 2633.10001
      C40.5761 0.3102-0.7563 3767.80.7970-0.0079 0.6039-2776.40.1814-0.9506-0.2518 2838.40001
    • Table 3. Number of iterations when convergence condition is reached

      View table

      Table 3. Number of iterations when convergence condition is reached

      Experimental sequenceAlgorithm of reference [14]Weighted LM algorithm
      1148
      21612
      31311
      41710
      5169
      6119
      71211
      81510
      91913
      101210
      Average14.510.3
    Tools

    Get Citation

    Copy Citation Text

    Aiguo Zhou, Jiangyang Yu, Jiakun Pu, Zhenyu Li, Ronghui Wei. Positioning Algorithm of Homonymous Object Points in Multi-Vision System Based on Weighted Levenberg-Marquardt[J]. Laser & Optoelectronics Progress, 2021, 58(24): 2415006

    Download Citation

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

    Category: Machine Vision

    Received: Jan. 29, 2021

    Accepted: Mar. 5, 2021

    Published Online: Dec. 1, 2021

    The Author Email: Jiangyang Yu (1930217@tongji.edu.cn), Zhenyu Li (zhenyu.li@tongji.edu.cn)

    DOI:10.3788/LOP202158.2415006

    Topics