Multi-level filter network for low-overlap point cloud registration

被引：0
|
作者：
He M. [1 ,2 ]
Liu L. [1 ,2 ]
Li S. [1 ,2 ]
Wu H. [1 ,2 ]
Zhu D. [1 ,2 ]
机构：
[1] Hubei Key Laboratory of Advanced Automotive Components Technology, Wuhan University of Technology, Wuhan
[2] Hubei Collaborative Innovation Center for Automotive Components Technology, Wuhan University of Technology, Wuhan
来源：
Guangxue Jingmi Gongcheng/Optics and Precision Engineering | 2024年 / 32卷 / 11期
关键词：
low-overlap point cloud; matching distortion; multi-level filter; partial measurement; point cloud registration;
D O I：
10.37188/OPE.20243211.1759
中图分类号：
学科分类号：
摘要：
Aiming at the problem of matching distortion caused by structural occlusion，field of view constraints，and stitching errors during point cloud reconstructed，a multi-level filter network（MulFNet）is proposed to achieve single-shot scanning point clouds for low-overlap registration. Firstly，the multi-level features of the point clouds are extracted through the feature pyramid coding network to obtain semantic information at different scales，and the attention module and the location module are embedded to enhance the feature significance；secondly，the multi-level features are filtered based on the multi-scale consistency voting mechanism，outliers are screened out and prominent features of the point clouds are retained to obtain the initial correspondence；and finally，the initial corresponding nodes are adaptively grouped based on the geometric relationships，and weighted estimation conversion is performed from local to global to obtain a prediction matrix based on the multi-level filtering. The experimental results show that the MulFNet is better than the popular networks such as FCGF and PREDATOR on the standard 3DMatch. The registration accuracy of the MulFNet on the scanning dataset with an average overlap rate of 10% is 40. 9% and 85. 4% higher than the ICP and the GeoTransformer，respectively. It is verified that the proposed network can effectively solve the problem of low-overlap point cloud matching distortion. © 2024 Chinese Academy of Sciences. All rights reserved.
引用
收藏
页码：1759 / 1772
页数：13
相关论文
共 35 条
  • [1] LIN S, ZHANG Q., Point cloud registration using neighborhood point information description and matching［J］, Opt. Precision Eng, 30, 8, pp. 984-997, (2022)
  • [2] YIN M，, ZHU Y Y，, YIN G F，, Et al., Deep feature interaction network for point cloud registration，with applications to optical measurement of blade profiles ［J］, IEEE Transactions on Industrial Informatics, 19, 8, pp. 8614-8624, (2023)
  • [3] MCKAY N D., A method for registration of 3-D shapes［J］, IEEE Transactions on Pattern Analysis and Machine Intelligence, 14, 2, pp. 239-256, (1992)
  • [4] DERBY-SHEV D Y，, Et al., AA-ICP：iterative closest point with Anderson acceleration［C］, 2018 IEEE International Conference on Robotics and Automation （ICRA）, pp. 3407-3412, (2018)
  • [5] YU Y W, WANG K, DU L Q，, Et al., Matching point pair optimization registration method for point cloud model［J］, Opt. Precision Eng, 31, 4, pp. 503-516, (2023)
  • [6] LIU Y SH, WU L, Et al., Mixed sparse iterative nearest point registration［J］, Opt. Precision Eng, 29, 9, pp. 2255-2267, (2021)
  • [7] ZHANG J Y, DENG B L., Fast and robust iterative closest point［J］, IEEE Transactions on Pattern Analysis and Machine Intelligence, 44, 7, pp. 3450-3466, (2022)
  • [8] WANG Z J，, Et al., WPMAVM：weighted plus-and-minus allowance variance minimization algorithm for solving matching distortion［J］, Robotics and Computer-Integrated Manufacturing, 76, (2022)
  • [9] WU H, FENG X ZH, HUA L, Et al., Local-to-global registration method of large complex components based on a de-pseudo-weighted variance minimization algorithm［J］, Scientia Sinica（Technologica）, 54, 3, pp. 422-442, (2024)
  • [10] BOLLES R C., Random Sample Consensus： A Paradigm for Model Fitting with Applications to Image Analysis and Automated Cartography［M］, Readings in Computer Vision, pp. 726-740, (1987)
1234