Acta Optica Sinica, Volume. 37, Issue 8, 0810004(2017)

Pavement Crack Recognition Based on Aerial Image

Bo Wang1, Xia Wang1、*, Fei Chen1, Yuntao He2, Wenguang Li2, and Li Liu2
Author Affiliations
  • 1 Key Laboratory of Optoelectronic Imaging Technology and System, Ministry of Education, School of Optoelectronics, Beijing Institute of Technology, Beijing 100081, China
  • 2 School of Aerospace Engineering, Beijing Institute of Technology, Beijing 100081, China
  • show less
    Figures & Tables(8)
    Original image and reproduction results of two scenes. (a) Original image of complex scene; (b) original image of simple scene; (c) processing result of complex scene; (d) processing result of simple scene
    Flow chart of recognition algorithm of aerial pavement crack
    Schematic of region growing based on multi-directional fitting. (a) Pixel labeling; (b) growing pixel; (c) growing region; (d) multi-directional fitting region
    Results of road extraction. (a) Traditional region growing; (b) region growing based on multi-directional fitting
    Comparison of detection results. (a) Traditional edge detection; (b) edge detection based on significance analysis
    Display results of crack length. (a) Detection result of left road; (b) detection result of right road
    • Table 1. Analysis of classified accuracy of crack image and run time

      View table

      Table 1. Analysis of classified accuracy of crack image and run time

      InstanceImage numberActual numberof crack imageDetectable numberof crack imageClassified accuracy /%Run time /s
      Instance 11009888.844.3
      Instance 210000100.021.6
      Instance 3100121083.352.7
      Instance 410011981.849.1
    • Table 2. Analysis of detection results of crack image length in instance 1

      View table

      Table 2. Analysis of detection results of crack image length in instance 1

      Crack 1Crack 2Crack 3Crack 4Crack 5Crack 6Crack 7Crack 8Crack 9
      Whether recognize classificationYesYesYesYesYesYesNoYesYes
      Average width of crack /mm10.611.411.29.712.811.18.511.713.6
      Actual length of crack /cm127.1139.8144.5112.3151.1132.387.6129.2160.7
      Detectable length of crack /cm102.1117.6118.886.9131.6108.90110.6143.3
      Detectable accuracy /%80.384.182.277.487.182.3085.689.2
    Tools

    Get Citation

    Copy Citation Text

    Bo Wang, Xia Wang, Fei Chen, Yuntao He, Wenguang Li, Li Liu. Pavement Crack Recognition Based on Aerial Image[J]. Acta Optica Sinica, 2017, 37(8): 0810004

    Download Citation

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

    Category: Image Processing

    Received: Mar. 24, 2017

    Accepted: --

    Published Online: Sep. 7, 2018

    The Author Email: Xia Wang (angelniuniu@bit.edu.cn)

    DOI:10.3788/AOS201737.0810004

    Topics