Laser & Optoelectronics Progress, Volume. 62, Issue 6, 0637001(2025)

Three-Dimensional Object Detection Method for Complex Aircraft Maintenance Scenes Based on Prior Information

Jun Wu1、*, Long Jin2, Shuo Huang2, Jinsong Lian2, Jiusheng Chen2, and Runxia Guo2
Author Affiliations
  • 1College of Aeronautical Engineering, Civil Aviation University of China, Tianjin 300300, China
  • 2College of Electronic Information and Automation, Civil Aviation University of China, Tianjin 300300, China
  • show less
    Figures & Tables(17)
    Schematic diagram of sensor coordinate system conversion
    Schematic diagram of feature points selection
    Schematic diagram of three-dimensional inspection process for airfoil
    Hardware equipments and scenarios for experiment
    Schematic diagram of sensor time synchronization. (a) Before synchronization; (b) after synchronization
    Image and point cloud of calibration plate. (a) Image; (b) point cloud
    Point cloud projection effect after sensor calibration
    Preprocessing effect of point cloud
    Partial images of dataset
    Visualization of target detection results
    Visualization of detection results by different algorithms. (a) DBSCAN + conditonal Euclidean clustering[12]; (b) K-means[13]; (c) proposed algorithm
    Schematic diagram of different distances
    Average time-consuming statistics of each link for proposed algorithm
    • Table 1. Calibration results of intrinsic matrix and extrinsic matrix

      View table

      Table 1. Calibration results of intrinsic matrix and extrinsic matrix

      ParameterValue
      Intrinsic matrix1775.76201251.36601777.4931147.238001
      Extrinsic matrix0.998-0.048-0.0090.016-0.011-0.039-0.999-0.1180.0480.998-0.041-0.0450    0     0     1     
    • Table 2. Comparison results of depth

      View table

      Table 2. Comparison results of depth

      Detection resultValue
      Average error0.212
      Measured depth6.508
      Algorithm output depth6.720
    • Table 3. Comparison of performances for various algorithms

      View table

      Table 3. Comparison of performances for various algorithms

      AlgorithmIoU of wing /%NTPClustering precision /%Average time /ms
      DBSCAN+conditional Euclidean clustering1289.5634887.88169.57
      K-means1351.5218646.9735.14
      Proposed94.7637594.7042.96
    • Table 4. Comparison results of distance

      View table

      Table 4. Comparison results of distance

      DistanceMeasured depthAlgorithm output depthError
      a1.7511.7430.008
      b1.1101.1140.004
      c3.0123.0010.011
      d1.3151.3090.006
      e10.43610.4110.025
      f0.5060.5180.012
    Tools

    Get Citation

    Copy Citation Text

    Jun Wu, Long Jin, Shuo Huang, Jinsong Lian, Jiusheng Chen, Runxia Guo. Three-Dimensional Object Detection Method for Complex Aircraft Maintenance Scenes Based on Prior Information[J]. Laser & Optoelectronics Progress, 2025, 62(6): 0637001

    Download Citation

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

    Category: Digital Image Processing

    Received: Jun. 17, 2024

    Accepted: Jul. 29, 2024

    Published Online: Mar. 5, 2025

    The Author Email:

    DOI:10.3788/LOP241495

    CSTR:32186.14.LOP241495

    Topics