Improved 3-D Point Cloud Registration Algorithm with Oriented Bounding Box

doi:10.3778/j.issn.1002-8331.2010-0443

Abstract

Abstract: In order to improve the registration accuracy of the source point cloud and the template point cloud when the initial relative deflection angle is too large, an improved PointNetLK algorithm, PointNetLK-OBB, combined with oriented bounding box is proposed. In this algorithm, the oriented bounding box of 3D point cloud is used to represent the macro characteristics of source point cloud and template point cloud. Under the guidance of iterative closest point algorithm, the oriented bounding box of source point cloud and template point cloud is aligned, and mirror symmetry effect is produced between source point cloud and template point cloud. According to the fitting degree of source point cloud and template point cloud, the mirror symmetry plane is detected, and the optimal rotation and translation of source point cloud is obtained to complete the registration task of 3D point cloud. In order to verify the effectiveness of the algorithm, a comparative experiment is carried out on the public data set ModelNet40. The experimental results show that, compared with PointNetLK, PointNetLK-OBB improves the registration accuracy of the source point cloud and the template point cloud when the initial relative deflection angle is too large, and the sensitivity of the initial relative position between the source point cloud and the template point cloud is reduced. The innovation of this paper is to use PointNetLK to avoid the non-convex problem of traditional point cloud registration, and to avoid the local optimization problem in PointNetLK context with the help of the regularity of oriented bounding box.

Key words: 3D point cloud registration, oriented bounding box, PointNetLK-OBB, mirror symmetry

摘要： 为了提升源点云和模板点云在初始相对偏转角度过大时的配准精度，提出了一种结合方向包围框的改进 PointNetLK算法PointNetLK-OBB。该算法用三维点云的方向包围框表示源点云和模板点云的宏观特征，在最近点迭代算法的引导下，对齐源点云和模板点云的方向包围框，并在源点云和模板点云间产生镜面对称效应；根据源点云和模板点云的拟合度探测镜面对称的对称面，得到源点云自身的最佳旋转和平移，完成三维点云配准任务。为了验证算法的有效性，在公开数据集ModelNet40上进行对比实验，实验结果显示，PointNetLK-OBB与PointNetLK相比，提升了源点云和模板点云在初始相对偏转角度过大时的配准精度，对源点云和模板点云间的初始相对位置敏感度降低。创新在于，利用PointNetLK绕开传统点云配准的非凸问题，借助于方向包围框的规整性避开PointNetLK语境下出现的局部最优问题。

关键词: 三维点云配准, 方向包围框, PointNetLK-OBB, 镜面对称 , , ,

LAN Wenhao, LI Ning, TONG Qiang. Improved 3-D Point Cloud Registration Algorithm with Oriented Bounding Box[J]. Computer Engineering and Applications, 2022, 58(14): 177-184.

兰文昊, 李宁, 佟强. 结合方向包围框改进的三维点云配准算法[J]. 计算机工程与应用, 2022, 58(14): 177-184.

References

[1] CHOY C，PARK J，KOLTUN V.Fully convolutional geometric features[C]//Proceedings of the IEEE International Conference on Computer Vision，2019：8958-8966.
[2] ECKART B，KIM K，KAUTZ J.Fast and accurate point cloud registration using trees of gaussian mixtures[J].arXiv：1807.02587，2018.
[3] GOJCIC Z，ZHOU C，WEGNER J D，et al.The perfect match：3D point cloud matching with smoothed densities[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，2019：5545-5554.
[4] WANG Y，SOLOMON J M.Deep closest point：learning representations for point cloud registration[C]//Proceedings of the IEEE International Conference on Computer Vision，2019：3523-3532.
[5] BESL P J，MCKAY N D.A method for registration of 3-D shapes[J].IEEE Transactions on Pattern Analysis and Machine Intelligence，1992，14（2）：239-256.
[6] SEGAL A，HAEHNEL D，THRUN S.Generalized-ICP[C]//Robotics：Science and Systems，2009.
[7] BOUAZIZ S，TAGLIASACCHI A，PAULY M.Sparse iterative closest point[C]//Computer Graphics Forum.Oxford，UK：Blackwell Publishing Ltd，2013：113-123.
[8] BIBER P，STRA?ER W.The normal distributions transform：a new approach to laser scan matching[C]//Proceedings of 2003 IEEE/RSJ International Conference on Intelligent Robots and Systems（IROS 2003），2003：2743-2748.
[9] AGAMENNONI G，FONTANA S，SIEGWART R Y，et al.Point clouds registration with probabilistic data association[C]//2016 IEEE/RSJ International Conference on Intelligent Robots and Systems（IROS），2016：4092-4098.
[10] YANG H，SHI J，CARLONE L.Teaser：fast and certifiable point cloud registration[J].arXiv：2001.07715，2020.
[11] YANG J，LI H，CAMPBELL 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] ROSEN D M，CARLONE L，BANDEIRA A S，et al.SE-Sync：a certifiably correct algorithm for synchronization over the special Euclidean group[J].The International Journal of Robotics Research，2019，38（2/3）：95-125.
[13] MARON H，DYM N，KEZURER I，et al.Point registration via efficient convex relaxation[J].ACM Transactions on Graphics，2016，35（4）：1-12.
[14] IZATT G，DAI H，TEDRAKE R.Globally optimal object pose estimation in point clouds with mixed-integer programming[M]//Robotics research.Cham：Springer，2020：695-710.
[15] QI C R，YI L，SU H，et al.Pointnet++：deep hierarchical feature learning on point sets in a metric space[C]//Advances in Neural Information Processing Systems，2017：5099-5108.
[16] PHAN A V，LE NGUYEN M，NGUYEN Y L H，et al.DGCNN：a convolutional neural network over large-scale labeled graphs[J].Neural Networks，2018，108：533-543.
[17] WU W，QI Z，FUXIN L.Pointconv：deep convolutional networks on 3d point clouds[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，2019：9621-9630.
[18] AOKI Y，GOFORTH H，SRIVATSAN R A，et al.Pointnetlk：robust & efficient point cloud registration using pointnet[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，2019：7163-7172.
[19] QI C R，SU H，MO K，et al.Pointnet：deep learning on point sets for 3d classification and segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，2017：652-660.
[20] LUCAS B D，KANADE T.An iterative image registration technique with an application to stereo vision[C]//Proc 7th Intl Joint Conf on Artificial Intelligence，1981：674-679.
[21] SARODE V，LI X，GOFORTH H，et al.Pcrnet：point cloud registration network using pointnet encoding[J].arXiv：1908.07906，2019.
[22] GELFAND N，MITRA N J，GUIBAS L J，et al.Robust global registration[C]//Symposium on Geometry Processing，2005.
[23] RUSU R B，BLODOW N，BEETZ M.Fast point feature histograms（FPFH） for 3D registration[C]//2009 IEEE International Conference on Robotics and Automation，2009：3212-3217.
[24] ZENG A，SONG S，NIE?NER M，et al.3dmatch：learning local geometric descriptors from RGB-D reconstructions[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，2017：1802-1811.
[25] DENG H，BIRDAL T，ILIC S.Ppfnet：global context aware local features for robust 3d point matching[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，2018：195-205.
[26] DENG H，BIRDAL T，ILIC S.Ppf-foldnet：unsupervised learning of rotation invariant 3d local descriptors[C]//Proceedings of the European Conference on Computer Vision（ECCV），2018：602-618.
[27] SCHNEIDER P，EBERLY D H.Geometric tools for computer graphics[M].[S.l.]：Elsevier，2002.
[28] CHANG C T，GORISSEN B，MELCHIOR S.Fast oriented bounding box optimization on the rotation group SO[(3,?)][J].ACM Transactions on Graphics，2011，30（5）：1-16.
[29] WU Z，SONG S，KHOSLA A，et al.3d shapenets：a deep representation for volumetric shapes[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition，2015：1912-1920.