Abstract
Rigid 3D point cloud registration is a challenging problem due to noise, outliers, variations in initial positions, and limited amounts of overlap. Existing methods lack a simple mechanism to deal with missing correspondence and usually fail to align point clouds in the presence of massive outliers and missing correspondences. Aiming at the problem of outliers and missing correspondences in the partially overlapping point cloud, a probability re-weighted 3D point cloud registration algorithm based on the Gaussian mixture model (GMM) is proposed in this paper. Firstly, the correspondences between the 3D target and source point clouds are established by the GMM and uniform distribution. We show that the missing correspondences in the target point cloud can be handled by re-weighting the mixing proportion of GMM through a prior probability re-weighting strategy. Secondly, we propose a posteriori probability inference strategy to infer the outliers and their proportion in the source point cloud, where the potential outliers are removed when solving the GMM parameters. Thirdly, the objective function in the form of point-to-plane distance is introduced by calculating the normal direction in the vicinity of the weight-averaged target point, and then the point clouds with large plane structures are registered finely. Finally, the experiments are conducted on Stanford 3D Scanning data and real 3D scene data. The overall RMSE on the former is 0.40 mm with fitness of 0.775, and it is 5.32 mm with fitness of 0.608 on the latter. The evaluation results show that the proposed algorithm can enhance the fitness and reduce the RMSE of the rigid 3D point cloud registration and improve the accuracy of registration.
Similar content being viewed by others
Availability of data and material
Not applicable.
Code availability
Not applicable.
References
Adams A, Baek J, Davis MA (2010) Fast high-dimensional filtering using the permutohedral lattice. In: Computer Graphics Forum, pp 753–762
Besl PJ, McKay ND (1992) Method for registration of 3-D shapes. In: sensor fusion IV: control paradigms and data structures, pp 586–606
Campbell D, Petersson L (2015) An adaptive data representation for robust point-set registration and merging. In: Proceedings of the IEEE International Conference on Computer Vision, pp 4292–4300
Chen Y, Medioni G (1992) Object modelling by registration of multiple range images. Image Vis Comput 10(3):145–155
Choi S, Zhou Q-Y, Koltun V (2015) Robust reconstruction of indoor scenes. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp 5556–5565
Dempster AP, Laird NM, Rubin DB (1977) Maximum likelihood from incomplete data via the EM algorithm. J R Stat Soc Ser B 39(1):1–22
Eckart B, Kim K, Kautz J (2018) Hgmr: hierarchical gaussian mixtures for adaptive 3d registration. In: proceedings of the European conference on computer vision (ECCV), pp 705–721
Eckart B, Kim K, Troccoli A, Kelly A, and Kautz J (2015) MLMD: maximum likelihood mixture decoupling for fast and accurate point cloud registration. In: international conference on 3d vision, pp 241–249
Evangelidis GD, Kounades-Bastian D, Horaud R, Psarakis EZ (2014) A generative model for the joint registration of multiple point sets. Comput Vis ECCV 2014:109–122
Gao W, Tedrake R (2019) Filterreg: robust and efficient probabilistic point-set registration using gaussian filter and twist parameterization. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp 11095–11104
Granger S, Pennec X (2002) Multi-scale EM-ICP: a fast and robust approach for surface registration. Comput Vis ECCV 2002:418–432
Handa A, Whelan T, McDonald J, Davison AJ (2014) A benchmark for RGB-D visual odometry, 3D reconstruction and SLAM. In: 2014 IEEE international conference on robotics and automation (ICRA), pp 1524–1531
Hartley R, Trumpf J, Dai Y, Li H (2013) Rotation averaging. Int J Comput Vis 103(3):267–305
Hertz A, Hanocka R, Giryes R, Cohen-Or D (2020) PointGMM: a neural GMM network for point clouds. In: 2020 IEEE/CVF conference on computer vision and pattern recognition (CVPR), pp 12051–12060
Hirose O (2020) A Bayesian formulation of coherent point drift. IEEE transactions on pattern analysis and machine intelligence 1-1
Hua B, Pham Q, Nguyen DT, Tran M, Yu L, Yeung S (2016) SceneNN: a scene meshes dataset with aNNotations. In: 2016 fourth international conference on 3D vision (3DV), pp 92–101
Jian B, Vemuri BC (2011) Robust point set registration using gaussian mixture models. IEEE Trans Pattern Anal Mach Intell 33(8):1633–1645
Li Y, Liu Y, Sun R, Guo R, Zhu L, Qi Y (2020) Multi-view point cloud registration with adaptive convergence threshold and its application in 3D model retrieval. Multimed Tools Appl 79(21):14793–14810
Mur-Artal R, Tardós JD (2017) ORB-SLAM2: an open-source SLAM system for monocular, stereo, and RGB-D cameras. IEEE Trans Robot 33(5):1255–1262
Myronenko A, Song X (2010) Point set registration: coherent point drift. IEEE Trans Pattern Anal Mach Intell 32(12):2262–2275
Newcombe RA, Fox D, Seitz SM (2015) Dynamic fusion: reconstruction and tracking of non-rigid scenes in real-time. In: 2015 IEEE conference on computer vision and pattern recognition (CVPR), pp 343–352
Pomerleau F, Colas F, Siegwart R (2015) A review of point cloud registration algorithms for mobile robotics. Foundations and trends® in Robotics 4:1–104
Pomerleau F, Liu M, Colas F, Siegwart R (2012) Challenging data sets for point cloud registration algorithms. Int J Robot Res 31(14):1705–1711
Pujol-Miró A, Casas JR, Ruiz-Hidalgo J (2019) Correspondence matching in unorganized 3D point clouds using convolutional neural networks. Image Vis Comput 83-84:51–60
Segal A, Hhnel D, Thrun S (2009) Generalized-ICP. Robot Sci Syst V 2(4)
Stoyanov T, Magnusson M, Lilienthal AJ (2013) Point set registration through minimization of the L2 distance between 3D-NDT models. In: IEEE International Conference on Robotics & Automation, pp 5196–5201
Tsin Y, Kanade T (2004) A correlation-based approach to robust point set registration. Comput Vis ECCV 2004:558–569
Whelan T, Salas-Moreno R, Glocker B, Davison A, Leutenegger S (2016) ElasticFusion: real-time dense SLAM and light source estimation. Int J Robot Res 35(14):1697–1716
Yuan W, Eckart B, Kim K, Jampani V, Kautz J (2020) DeepGMR: learning latent Gaussian mixture models for registration. Comput Vis ECCV 2020:733–750
Zhou J, Ma X, Liang L, Yuhe L, Yang Y (2019) Robust variational Bayesian point set registration. In: 2019 IEEE/CVF international conference on computer vision (ICCV), pp 9904–9913
Zhou Q-Y, Park J, Koltun V (2018) Open3D: a modern library for 3D data processing. CoRR abs/1801.09847
Zhou Z, Tu J, Geng C, Hu J, Tong B, Ji J, Dai Y (2018) Accurate and robust non-rigid point set registration using Student’s-t mixture model with prior probability modeling. Sci Rep 8(1):8742
Acknowledgements
This work was supported partially by the National Natural Science Foundation of China (No.51875152), the Graduate Students Education Innovation Foundation of Shanxi Province (No.2021Y698) and the Natural Science Foundation of Shanxi Province (No.201801D121134).
Funding
This work was supported partially by the National Natural Science Foundation of China (No.51875152), the Graduate Students Education Innovation Foundation of Shanxi Province (No.2021Y698) and the Natural Science Foundation of Shanxi Province (No.201801D121134).
Author information
Authors and Affiliations
Contributions
Zhiliang Sun: Conceptualization, Methodology, Software, Data curation, Validation, Visualization, Writing-Original draft preparation.
Rongguo Zhang: Project administration, Supervision, Writing-Reviewing and Editing, Investigation, Funding acquisition.
Jing Hu: Formal analysis, Software, Supervision.
Xiaojun Liu: Methodology, Funding acquisition.
Corresponding author
Ethics declarations
Ethics approval
Not applicable.
Consent to participate
All authors each made a significant contribution to the research reported and have read and approved the submitted manuscript.
Consent for publication
The manuscript is approved by all authors for publication. The authors claim that none of the material in the paper has been published or is under consideration for publication elsewhere.
Conflicts of interest/competing interests
The authors declare that there are no competing interests regarding the publication of this paper.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Sun, Z., Zhang, R., Hu, J. et al. Probability re-weighted 3D point cloud registration for missing correspondences. Multimed Tools Appl 81, 11107–11126 (2022). https://doi.org/10.1007/s11042-022-12134-5
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-022-12134-5