• Acta Optica Sinica
  • Vol. 40, Issue 21, 2110001 (2020)
Guiping Cao, Xingsi Liu, Nian Liu, Kecheng Yang, and Min Xia*
Author Affiliations
  • School of Optical and Electronic Information, Huazhong University of Science and Technology, Wuhan, Hubei 430074, China
  • show less
    DOI: 10.3788/AOS202040.2110001 Cite this Article Set citation alerts
    Guiping Cao, Xingsi Liu, Nian Liu, Kecheng Yang, Min Xia. Segmentation of Subway Tunnel Wall Surface Objects Based on Laser 3D Point Cloud[J]. Acta Optica Sinica, 2020, 40(21): 2110001 Copy Citation Text show less
    System block diagram of the detection system
    Fig. 1. System block diagram of the detection system
    Photo of the detection system
    Fig. 2. Photo of the detection system
    Schematic diagram of polar coordinates
    Fig. 3. Schematic diagram of polar coordinates
    A picture of a simulated tunnel
    Fig. 4. A picture of a simulated tunnel
    Three-dimensional point cloud image of a tunnel
    Fig. 5. Three-dimensional point cloud image of a tunnel
    Raw point cloud data
    Fig. 6. Raw point cloud data
    Point cloud data after segmentation
    Fig. 7. Point cloud data after segmentation
    Schematic for defining concept of DBSCAN
    Fig. 8. Schematic for defining concept of DBSCAN
    Flowchart of segmentation algorithm based on density clustering
    Fig. 9. Flowchart of segmentation algorithm based on density clustering
    Segmentation results based on density clustering algorithm
    Fig. 10. Segmentation results based on density clustering algorithm
    Scene 1
    Fig. 11. Scene 1
    Scene 2
    Fig. 12. Scene 2
    3D point cloud of scene 1
    Fig. 13. 3D point cloud of scene 1
    3D point cloud of scene 2
    Fig. 14. 3D point cloud of scene 2
    Result of scene 1. (a) Result of region growing segmentation; (b) result based on density clustering segmentation method
    Fig. 15. Result of scene 1. (a) Result of region growing segmentation; (b) result based on density clustering segmentation method
    Result of scene 2. (a) Result of region growing segmentation; (b) result based on density clustering segmentation method
    Fig. 16. Result of scene 2. (a) Result of region growing segmentation; (b) result based on density clustering segmentation method
    Deformation of the tunnel is simulated by object compression
    Fig. 17. Deformation of the tunnel is simulated by object compression
    A deformed simulated tunnel
    Fig. 18. A deformed simulated tunnel
    Segmentation results of object point cloud in the deformed tunnel. (a) Result of region growing segmentation; (b) result based on density clustering segmentation method
    Fig. 19. Segmentation results of object point cloud in the deformed tunnel. (a) Result of region growing segmentation; (b) result based on density clustering segmentation method
    Object and its positionLength /mmWidth /mmHeight /mm
    PVC water pipe (left in the Fig. 6)580110110
    Carton 1 (left in the Fig. 6)18513276
    Carton 2,3,4 (right in the Fig. 6)20715770
    Table 1. Parameters of objects for testing region growing segmentation
    Object and its positionLength /mmWidth /mmHeight /mm
    PVC water pipe (left in the Fig. 14)580110110
    Aluminum tube 1 (left in the Fig. 14)5005050
    Aluminum tube 2 (above in the Fig. 14)5005050
    Table 2. Parameters of objects in scene 1
    Object and its positionLength /mmWidth /mmHeight /mm
    PVC water pipe (left in the Fig. 15)580110110
    Aluminum tube (above in the Fig. 15)5005050
    Paper tube (right in the Fig. 15)13507070
    Carton 1 (left in the Fig. 15)390260170
    Carton 2 (right in the Fig. 15)41045080
    Table 3. Parameters of objects in scene 2
    Guiping Cao, Xingsi Liu, Nian Liu, Kecheng Yang, Min Xia. Segmentation of Subway Tunnel Wall Surface Objects Based on Laser 3D Point Cloud[J]. Acta Optica Sinica, 2020, 40(21): 2110001
    Download Citation