计算机科学

计算机科学 ›› 2024, Vol. 51 ›› Issue (2): 142-150.doi: 10.11896/jsjkx.230200073

• 计算机图形学&多媒体 • 上一篇    下一篇

结合注意力机制的多重引导点云配准网络

刘旭珩, 柏正尧, 许祝, 杜佳锦, 肖霄   

  1. 云南大学信息学院 昆明650221
  • 收稿日期:2023-02-20 修回日期:2023-06-20 出版日期:2024-02-15 发布日期:2024-02-22
  • 通讯作者: 柏正尧(baizhy@ynu.edu.cn)
  • 作者简介:(liuxuheng@mail.ynu.edu.cn)
  • 基金资助:
    云南省重大科技专项课题(202002AD080001);云南大学第十四届研究生科研创新项目(KC-22222543)

Multi-guided Point Cloud Registration Network Combined with Attention Mechanism

LIU Xuheng, BAI Zhengyao, XU Zhu, DU Jiajin, XIAO Xiao   

  1. School of Information Science and Engineering,Yunnan University,Kunming 650221,China
  • Received:2023-02-20 Revised:2023-06-20 Online:2024-02-15 Published:2024-02-22
  • About author:LIU Xuheng,born in 1998,postgra-duate.His main research interests include point cloud registration and three-dimensional reconstruction.BAI Zhengyao,born in 1967,Ph.D,professor,master supervisor.His main research interests include signal proces-sing,image processing,pattern recognition and machine learning,etc.
  • Supported by:
    Yunnan Provincial Major Science and Technology Special Plan Projects(202002AD080001) and 14th Postgra-duate Research Innovation Project of Yunnan University(KC-22222543).

PDF (PC)

1692

摘要： 针对点云配准过程中仅仅利用点云特征寻求对应关系使得离群点多、配准精度不高的问题进行研究,提出了一种使用点云之间匹配点概率矩阵和点云空间信息特征矩阵共同搜寻对应关系,并且相互配合确定对应点权重的点云配准网络——AMGNet。首先使用点云特征提取网络获得两片待配准点云的高维特征;然后采用Transformer对独立特征进行上下文信息融合,之后利用关键点提取模块选取出特征更强的点,使用SoftBBS方法获得点云匹配点概率矩阵后,结合点云空间特征矩阵搜索到最终的对应关系,同时,权重分配也使用了双重矩阵共同决定的策略;最后使用奇异值分解获得需要的刚性变换矩阵。在ModelNet40,7Scenes等人工合成数据集和真实场景数据集上进行了多次实验。结果表明,在ModelNet40目标未知实验中的旋转矩阵和平移向量的均方误差分别降低至0.025和0.004 6。AMGNet配准精度较高,抗干扰能力强,泛化能力强。

关键词: 点云配准, 注意力机制, 多重矩阵引导, 加权SVD

Abstract: This paper proposes a point cloud alignment network,AMGNet,which uses the probability matrix of matching points between point clouds and the spatial information feature matrix of point clouds to search for correspondence and determine the weights of corresponding points with each other.First,the point cloud feature extraction network is used to get the high-dimensional features of the two unaligned point clouds and then the Transformer is used to fuse the independent features with the contextual information.Also,the weight assignment uses the strategy of double matrix co-determination.Finally,the singular value decomposition is used to obtain the required rigid transformation matrix.Several experiments are conducted on synthetic datasets,such as ModelNet40,7Scenes and real scenes.The results show that the mean square error of rotation matrix and translation vector in ModelNet40 target unknown experiments is reduced to 0.025 and 0.004 6,respectively.AMGNet alignment has high accuracy,high interference resistance,and good generalization ability.

Key words: Point cloud registration, Attention mechanism, Multiple matrix guidance, Weighted SVD

中图分类号: 

  • TP391.41
[1]LI Z M,ZHANG Y P,LIU Y J,et al.Deformable Graph Convolutional Networks Based Point Cloud Representation Learning[J].Computer Science,2022,49(8):273-278.
[2]HUANG X,MEI G,ZHANG J,et al.A comprehensive survey on point cloud registration[J].arXiv:2103.02690,2021.
[3]QIN H X,LIU Z T,TAN B Y.Review on deep learning rigid point cloud registration[J].Journal of Image and Graphics,2022,27(2):329-348.
[4]LI J W,ZHAN J W.Review on 3D point cloud registrationmethod[J].Journal of Image and Graphics,2022,27(2):349-367.
[5]LI J,ZHANG C,XU Z,et al.Iterative distance-aware similarity matrix convolution with mutual-supervised point elimination for efficient point cloud registration[C]//European Conference on Computer Vision.Berlin:Springer,2020:378-394.
[6]WANG H,LIU X,KANG W,et al.Multi-features guidance net-work for partial-to-partial point cloud registration[J].Neural Computing and Applications,2022,34(2):1623-1634.
[7]HEZRONI I,DRORY A,GIRYES R,et al.DeepBBS:Deep Best Buddies for Point Cloud Registration[C]//2021 International Conference on 3DVision(3DV).Piscataway,NJ:IEEE.2021:342-351.
[8]WANG Y,SOLOMON J M.Deep closest point:Learning rep resentations for point cloud registration[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.Piscataway,NJ:IEEE 2019:3523-3532.
[9]YEW Z J,LEE G H.RPM-Net:Robust Point Matching usingLearned Features[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR).Piscataway,NJ:IEEE,2020:11821-11830.
[10]BESL P J,MCKAY N D.Method for registration of 3-D shapes[C]//Sensor Fusion IV:Control Paradigms and Data Structures.Spie,1992:586-606.
[11]YANG J,LI H,CAMPBALL D,et al.Go-ICP:A globally optimal solution to 3D ICP point-set registration[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2015,38(11):2241-2254.
[12]HEXSEL B,VHAVLE H,CHEN Y.DICP:Doppler IterativeClosest Point Algorithm[J].arXiv:2201.11944,2022.
[13]MAGNUSSON M,LILIENTHAL A,DUCKETT T.Scan registration for autonomous mining vehicles using 3D-NDT[J].Journal of Field Robotics,2007,24(10):803-827.
[14]AIGER D,MITRA N J,COHEN-OR D.4-points congruent sets for robust pairwise surface registration[J].ACM Transactions on Graphics,2008,27(3):1-10.
[15]FISCHLER M A,BOLLES R C.Random sample consensus:aparadigm for model fitting with applications to image analysis and automated cartography[J].Communications of the ACM,1981,24(6):381-395.
[16]MELLADO N,AIGER D,MITRA N J.Super 4PCS Fast Global Pointcloud Registration via Smart Indexing[J].Computer Graphics Forum,2014,33(5):205-215.
[17]KAMOUSI P,LAZARD S,MAHESHWARI A,et al.Analysis of farthest point sampling for approximating geodesics in a graph[J].Computational Geometry,2016,57:1-7.
[18] RUSINKIEWICZ S,LEVOV M.Efficient variants of the ICP algorithm[C]//Proceedings third International Conference on 3-D Digital Imaging and Modeling.Piscataway,NJ:IEEE,2001:145-152.
[19]PRAKHYA S M,LIU B,LIN W.Detecting keypoint sets on 3D point clouds via Histogram of Normal Orientations[J].Pattern Recognition Letters,2016,83:42-48.
[20]CHUA C S.Point Signatures:A New Representation for 3D Object Recognition[J].International Journal of Computer Vision,1997,25:63-85.
[21]FROME A,HUBER D,KOLLURI R,et al.Recognizing objects in range data using regional point dscriptors[C]//European Conference on Computer Vision.Berlin:Springer:2004:224-237.
[22]RUSU R B,BLODOW N,MARTON Z C,et al.Aligning point cloud views using persistent feature histograms[C]//2008 IEEE/RSJ International Conference on Intelligent Robots and Systems.Piscataway,NJ:IEEE.2008:3384-3391.
[23]RUSU R B,BLODOW N,BEETZ M.Fast point feature histo-gra.ms(FPFH) for 3D registration[C]//2009 IEEE International Conference on Robotics and Automation.Piscataway,NJ:IEEE,2009:3212-3217.
[24] ZHANG W L,CHENG L,REN M F,et al.Point Cloud Registration Based on AGConv Local Feature Descriptors[J].Computer Engineering,2023,49(11):231-237.
[25] LI X M,WANG C Y,LIU Xl,et al.Point cloud registration method based on supervoxel bidirectional nearest neighbor distance ratio[J].Journal of Jilin University(Engineering and Technology Edition),2022,52(8):1918-1925.
[26]AOKI Y,GOFORTH H,SRIVATSAN R A,et al.Pointnetlk:Robust & efficient point cloud registration using pointnet[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Piscataway,NJ:IEEE,2019:7163-7172.
[27]LU W,WAN G,ZHOU Y,et al.Deepvcp:An end-to-end deep neural network for point cloud registration[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision.Piscataway,NJ:IEEE,2019:12-21.
[28]MONTI F,BOSCAINI D,MASCI J,et al.Geometricdeep lear-ning on graphs and manifolds using mixture model cnns[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition(CVPR).Piscataway,NJ:IEEE,2017:5425-5434.
[29]CHARLES R Q,SU H,KAICHUN M,et al.PointNet:Deep Learning on Point Sets for 3D Classification and Segmentation[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition(CVPR).Piscataway,NJ:IEEE,2017:77-85.
[30]WANG Y,SUN Y,LIU Z,et al.Dynamic graph cnn for learning on point clouds[J].ACM Transactions on Graphics(TOG),2019,38(5):1-12.
[31]QI C R,YI L,SU H,et al.Pointnet++:Deep hierarchical feature learning on point sets in a metric space[J].arXiv:1706.02413,2017.
[32]VASWANI A,SHAZEER N,PARMAR N,et al.Attention isall you need[J].arXiv:1706.03762,2017.
[33]LI J,ZHANG C,XU Z,et al.Iterative distance-aware similarity matrix convolution with mutual-supervised point elimination for efficient point cloud registration[C]//Computer Vision-ECCV 2020:16th European Conference,Part XXIV 16.Springer International Publishing,2020:378-394.
[34]WU Z R,SONG S,KHOSLA A,et al.3D ShapeNets:A deep representation for volumetric shapes[C]//2015 IEEE Confe-rence on Computer Vision and Pattern Recognition(CVPR).Piscataway,NJ:IEEE,2015:1912-1920.
[35] SHOTTON J,GLOCKER B,ZACH C,et al.Scene coordinate regression forests for camera relocalization in RGB-D images[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Piscataway,NJ:IEEE,2013:2930-2937.
[36] ZHOU Q Y,PARK J,KOLTUN V.Open3D:A modern library for 3D data processing[J].arXiv:1801.09847,2018.
[37]SHEN Y,HUI L,JIANG H,et al.Reliable Inlier Evaluation for Unsupervised Point Cloud Registration[C]//Proceedings of the AAAI Conference on Artificial Intelligence.California:Association for the Advancement of Artificial Intelligence(AAAI),2022:2198-2206.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发