IMPROVED DEEP LEARNING-GUIDED SPARSE ICP FOR POINT CLOUDS REGISTRATION IN RAIL WEAR CALCULATION

Xueyin Liu,∗,∗∗ Chen Yan,∗ Yao Fu,∗∗∗ and Peng Chen∗

References

  1. [1] C. Ngamkhanong, S. Kaewunruen, and B.J.A. Costa, State-of-the-art review of railway track resilience monitoring,Infrastructures, 3(1), 2018, 3.15
  2. [2] S. Wang, B. Wen, D. Ren, Y. Ding, Z. Wen, D. Dong, and G.Tao, Analysis of the causes of severe side wear of the high rail onmetro curves by numerical simulation and field investigation,Proceedings of the Institution of Mechanical Engineers, PartF: Journal of Rail and Rapid Transit, 238(4), 2024, 381–393.
  3. [3] H. Liu, Y. Li, Z. Ma, and C. Wang, Recognition and calibrationof rail profile under affine- distortion-based point set mapping,IEEE Transactions on Instrumentation and Measurement,66(1), 2016, 131–140.
  4. [4] W. Jin, X. Zhan, and B. Jiang, Non-contact rail-wear inspectingsystem based on image understanding, Proceeding of theInternational Conference Mechatronics and Automation, 2007,3854–3858.
  5. [5] S. Cai, G. Bao, and J. Pang, A structured light-basedvisual sensing system for detecting multi-layer and multi-trackwelding, International Journal of Robotics and Automation,36(4), 2001, 264–273.
  6. [6] D. Zhang, S. N. Lingamanaik, and H. Chung, Image-based 3dreconstruction for rail profile measurement, Proceedings of theInstitution of Mechanical Engineers, Part-F: Journal of Railand Rapid Transit, 237(3), 2023, 309–321.
  7. [7] P.J. Besl and N.D. McKay, Method for registration of 3-D shapes, Sensor Fusion IV: Control Paradigms and DataStructures, 1611 1992, 586–606.
  8. [8] O. Ervan and H. Temeltas, A histogram-based sampling methodfor point cloud registration, The Photogrammetric Record,38(183), 2023, 210–232.
  9. [9] C.R. Qi, L. Yi, H. Su, and L.J. Guibas, Pointnet++: Deephierarchical feature learning on point sets in a metric space,Proceeding of the 31st Advances in Neural InformationProcessing Systems, 2017, 5105–5114.
  10. [10] A. Mahdaoui, A. Bouazi, A. Marhraoui Hsaini, and E.Sbai, Entropic method for 3D point cloud simplification, inProcceeding of the Mediterranean Symposium on Smart CityApplications, 2017, 613–621.
  11. [11] S. Chen, D. Tian, C. Feng, A. Vetro, and J. Kovacevic’, Fastresampling of three-dimensional point clouds via graphs, IEEETransactions on Signal Processing, 66(3), 2017, 666–681.
  12. [12] S. Bouaziz, A. Tagliasacchi, and M. Pauly, Sparse iterativeclosest point, Computer Graphics Forum, 32, 2013, 113–123.
  13. [13] S. Irani and P. Raghavan, Combinatorial and experimentalresults for randomized point matching algorithms, in Proceedingof the . 12th Annual Symposium on Computational Geometry,1996, 68–77.
  14. [14] C. Ji, Y. Li, J. Fan, and S. Lan, A novel simplification methodfor 3D geometric point cloud based on the importance of point,IEEE Access, 7, 2019, 129 029–129 042.
  15. [15] Y. Wang and L. Zhao, Point cloud sampling method based onoffset-attention and mutual supervision, The Visual Computer,39(6), 2013, 2337–2345.
  16. [16] Z. Wu, S. Song, A. Khosla, F. Yu, L. Zhang, X. Tang, andJ. Xiao, 3D shapenets: A deep representation for volumetricshapes, in Proceedings of the IEEE Conference on ComputerVision and Pattern Recognition (CVPR), 2015, 1912–1920.
  17. [17] W. Lu, G. Wan, Y. Zhou, X. Fu, and S. Song, Deepicp: Anend-to-end deep neural network for 3D point cloud registration.2019, arXiv:1905.04153.
  18. [18] Z. Gojcic, C. Zhou, J.D. Wegner, and A. Wieser, The perfectmatch: 3D point cloud matching with smoothed densities, inProceeding of the IEEE/CVF Conference on Computer Visionand Pattern Recognition (CVPR), 2018, 5540–5549.
  19. [19] Z. Gojcic, C. Zhou, J. D. Wegner, L.J. Guibas, and T. Birdal,Learning multiview 3D point cloud registration, in Proceedingsof the IEEE/CVF Conference on Computer Vision and PatternRecognition, 2020, 1759–1769.
  20. [20] A. Zeng, S. Song, M. Nießner, M. Fisher, J. Xiao, and T.Funkhouser, 3Dmatch: Learning local geometric descriptorsfrom RGB-D reconstructions, in Proceeding IEEE Conf. onComputer Vision and Pattern Recognition, 2017, 1802–1811.
  21. [21] K. Li, Y. Dong, B. Park, T. Koch, and Z. Wan, POE-Net: Expand-Laplacian attention network for large-scale placerecognition in point cloud, International Journal of Roboticsand Automation, 206, 2025, 222–229.
  22. [22] J. Bae and S. Han, Vision-based inspection approach using aprojector-camera system for off-site quality control in modularconstruction: Experimental investigation on operational condi-tions, Journal of Computing in Civil Engineering, 35(5), 2021,04021012.
  23. [23] L. Baofeng, T. Baohua, Z. Xiaoming, and Z. Mingling,Accelerator of the global automated image registrationalgorithm, Journal of Computer-Aided Design & ComputerGraphics, 24(10), 2012, 1363–1368.
  24. [24] X. Zhang, A. Zhang, and Z. Wang, Point cloud registrationbased on improved normal distribution transform algorithm,Laser & Optoelectronics Progress, 51(4), 2014, 96–105.
  25. [25] X. Zhang, A. Zhang, and Z. Wang, Point cloud registrationbased on improved normal distribution transform algorithm,Laser & Optoelectronics Progress, 51(4), 2014, 96–105 Laser& Optoelectronics Progress, 51(4), 2014, 96–105.
  26. [26] Y. Diez, F. Roure, X. Llado’, and J. Salvi, A qualitative reviewon 3d coarse registration methods, ACM Computing Surveys(CSUR), 47(3), 2015, 1–36.
  27. [27] J. Yang, H. Li, D. Campbell, and Y. Jia, Go-ICP: A globallyoptimal solution to 3D ICP pointset registration, IEEETransactions on Pattern Analysis and Machine Intelligence,38(11), 2015, 2241–2254.
  28. [28] P. Nousi, A. Tefas, and I. Pitas, Dense convolutional featurehistograms for robust visual object tracking, Image and VisionComputing, 99, 2020, 103933.
  29. [29] Y. Liu, D. Kong, D. Zhao, X. Gong, and G. Han, A pointcloud registration algorithm based on feature extraction andmatching, Mathematical Problems in Engineering, 2018(1),2018, 7352691.
  30. [30] T. Le and M. Cuturi, Adaptive euclidean maps for histograms:generalised Aitchison embeddings, Machine Learning, 99, 2015,169–187.
  31. [31] C. Aguilera, F. Barrera, F. Lumbreras, A.D. Sappa, and R.Toledo, Multispectral image feature points, Sensors, 12(9),2012, 12661–12672.
  32. [32] R.B. Rusu, N. Blodow, and M. Beetz, Fast point featurehistograms (FPFH) for 3D registration, in Proceeding IEEEInternational Conference on Robotics and Automation, 2009,3212–3217.
  33. [33] Y. Aoki, H. Goforth, R.A. Srivatsan, and S. Lucey, Pointnetlk:Robust & efficient point cloud registration using pointnet, inProceeding IEEE/CVF Conference on Computer Vision andPattern Recognition (CVPR), 2019.
  34. [34] Y. Wang and J. Solomon, Deep closest point: Learningrepresentations for point cloud registration, in ProceedingIEEE/CVF International Conference on Computer Vision(ICCV), 2019, 3522–3531.
  35. [35] Z.J. Yew and G.H. Lee, RPM-Net: Robust point matchingusing learned features, in Proceeding Conference . on ComputerVision and Pattern Recognition (CVPR), 2020, 11821–11830.

Important Links:

Go Back