• Laser & Optoelectronics Progress
  • Vol. 62, Issue 2, 0237006 (2025)
Rufei Liu1,*, Weibin Xu1, Qianying Zhao2, and Zhanwen Su1
Author Affiliations
  • 1College of Geodesy and Geomatics, Shandong University of Science and Technology, Qingdao 266590, Shandong , China
  • 2Qingdao Geology and Minerals Rock and Soil Engineering Co., Ltd., Qingdao 266072, Shandong , China
  • show less
    DOI: 10.3788/LOP241089 Cite this Article Set citation alerts
    Rufei Liu, Weibin Xu, Qianying Zhao, Zhanwen Su. Adaptive Descent Distance Cloth Simulation Algorithm for Pavement Pothole Extraction[J]. Laser & Optoelectronics Progress, 2025, 62(2): 0237006 Copy Citation Text show less
    Flowchart of pavement pothole extraction process
    Fig. 1. Flowchart of pavement pothole extraction process
    Rendering images of the pavement point cloud. (a) True color rendering image; (b) elevation three-dimensional rendering image
    Fig. 2. Rendering images of the pavement point cloud. (a) True color rendering image; (b) elevation three-dimensional rendering image
    Simulated process of cloth drop. (a) Original simulated process of cloth drop; (b) improved simulated process of cloth drop
    Fig. 3. Simulated process of cloth drop. (a) Original simulated process of cloth drop; (b) improved simulated process of cloth drop
    Rendering images of depth information under different internal forces. (a) Internal force of 1; (b) internal force of 5; (c) internal force of 10
    Fig. 4. Rendering images of depth information under different internal forces. (a) Internal force of 1; (b) internal force of 5; (c) internal force of 10
    Elevation rendering image of measured data
    Fig. 5. Elevation rendering image of measured data
    Results of pothole detection
    Fig. 6. Results of pothole detection
    Panoramic and local magnification of detection results for Fig. 6. (a) Region A; (b) region B; (c) region C; (d) region D; (e) region E; (f) region F
    Fig. 7. Panoramic and local magnification of detection results for Fig. 6. (a) Region A; (b) region B; (c) region C; (d) region D; (e) region E; (f) region F
    Comparison of results by different algorithms. (a) Proposed algorithm; (b) traditonal CSF algorithm; (c) RANSAC; (d) cross-sectional comparison between the datum plane constructed by different algorithms and the pavement point cloud
    Fig. 8. Comparison of results by different algorithms. (a) Proposed algorithm; (b) traditonal CSF algorithm; (c) RANSAC; (d) cross-sectional comparison between the datum plane constructed by different algorithms and the pavement point cloud
    AlgorithmNTPNFPNFNR /%P /%
    Proposed355783.387.5
    Traditional CSF29111565.972.5
    RANSAC3317978.666.0
    Table 1. Detection results of potholes
    PotholeGround truth /m2Proposed algorithmTraditional CSF algorithmRANSAC algorithm
    Measure /m2AE /m2RE /%Measure /m2AE /m2RE /%Measure /m2AE /m2RE /%
    10.12430.1023-0.022017.6990.0865-0.037830.4100.0948-0.029523.733
    20.23170.1925-0.039216.9180.1395-0.092239.7930.1668-0.064928.010
    30.56260.5253-0.03736.6300.3743-0.188333.4700.4337-0.128922.911
    40.14020.15720.017012.1260.16370.023516.7620.1317-0.00856.063
    50.03590.0320-0.003910.8640.0238-0.012133.7050.0287-0.007220.056
    Table 2. Area extraction results of potholes
    PotholeGround truth /mProposed algorithmTraditional CSF algorithmRANSAC algorithm
    Measure /mAE /mRE /%Measure /mAE /mRE /%Measure /mAE /mRE /%
    10.02910.0273-0.00186.1860.0214-0.007726.4600.0239-0.005217.869
    20.03450.0329-0.00164.6380.0237-0.010831.3040.0278-0.006719.420
    30.03310.0299-0.00329.6680.0256-0.007522.6590.03870.005616.918
    40.02410.0224-0.00177.0540.02670.002610.7880.0194-0.004719.502
    50.02170.0196-0.00219.6770.0178-0.003917.9720.0172-0.004520.737
    Table 3. Depth extraction results of potholes