基于关键点精确配对的点云曲面匹配方法

doi:10.16182/j.issn1004731x.joss.22-0146

系统仿真学报 ›› 2023, Vol. 35 ›› Issue (6): 1169-1182.doi: 10.16182/j.issn1004731x.joss.22-0146

基于关键点精确配对的点云曲面匹配方法

宁小娟¹^,²(), 李春旭¹(), 王嘉豪¹, 唐婧¹, 王映辉³, 金海燕¹^,²

^1.西安理工大学计算机科学与工程学院，陕西西安 710048
^2.陕西省网络计算与安全技术重点实验室，陕西西安 710048
^3.江南大学人工智能与计算机学院，江苏无锡 214122

收稿日期:2022-02-08 修回日期:2022-04-13 出版日期:2023-06-29 发布日期:2023-06-20
通讯作者: 李春旭 E-mail:ningxiaojuan@xaut.edu.cn;876167524@qq.com
作者简介:宁小娟(1982-)，女，教授，博士，研究方向为模式识别与图像处理。Email：ningxiaojuan@xaut.edu.cn
基金资助:
国家自然科学基金(61871320);西安市碑林区科技计划(GX2107)

Point Cloud Surface Matching Method Based on Precise Matching of Critical Point

Xiaojuan Ning¹^,²(), Chunxu Li¹(), Jiahao Wang¹, Jing Tang¹, Yinghui Wang³, Haiyan Jin¹^,²

^1.School of Computer Science and Engineering, Xi'an University of Technology, Xi'an 710048, China
^2.Shaanxi Provincial Key Laboratory of Network Computing and Security Technology, Xi'an 710048, China
^3.School of Artificial Intelligence and Computer, Jiangnan University, Wuxi 214122, China

Received:2022-02-08 Revised:2022-04-13 Online:2023-06-29 Published:2023-06-20
Contact: Chunxu Li E-mail:ningxiaojuan@xaut.edu.cn;876167524@qq.com

摘要/Abstract

摘要：

针对基于特征的点云曲面匹配方法在关键点匹配时匹配效率低和精度不够的问题，提出了一种基于关键点精确配对的点云曲面匹配方法。通过采用基于曲率信息的改进3D-SIFT(3D scale-invariant feature transform)算法，提取点云数据的关键点；将关键点处的FPFH(fast point feature histograms)特征描述以及模型中心点到关键点的向量与模型主趋势的夹角作为约束条件,获取精确的关键点匹配点对集合；求解刚体变换参数实现模型曲面的初始匹配；使用ICP(iterative closest point)算法进行二次优化，实现模型曲面的精确匹配。实验表明：该方法既能解决关键点匹配精度问题，又能很好地解决匹配效率的问题。

关键词: 尺度不变特征变换, 快速点特征直方图, 主趋势, 刚体变换参数, 迭代最近点

Abstract:

To solve the low matching efficiency and insufficient accuracy of feature-based point cloud surface matching method during critical point matching, a point cloud surface matching method based on the pairing exaction of critical points is proposed. An improved 3D scale-invariant feature transform(3D-SIFT) algorithm based on curvature information is presented to extract the critical points. Fast point feature histograms(FPFH) feature, the angle between the vector from the center to critical points and the principal direction of the model are taken as the constraints to obtain the exact critical point matching point pair set. The initial matching of the model surface is implemented by the rigid body transformation parameters, and further the accurate matching of the model surface is achieved by iterative closest point(ICP). Experiments show that the approach can not only improve the critical point matching accuracy, but also enhance the matching efficiency. Compared with other methods, the method is slightly better on the matching speed.

Key words: scale-invariant feature transformation, fast point feature histogram, principal direction, rigid body transformation parameters, iterative closest point

中图分类号:

TP391

宁小娟, 李春旭, 王嘉豪, 唐婧, 王映辉, 金海燕. 基于关键点精确配对的点云曲面匹配方法[J]. 系统仿真学报, 2023, 35(6): 1169-1182.

Xiaojuan Ning, Chunxu Li, Jiahao Wang, Jing Tang, Yinghui Wang, Haiyan Jin. Point Cloud Surface Matching Method Based on Precise Matching of Critical Point[J]. Journal of System Simulation, 2023, 35(6): 1169-1182.

图/表 20

图1

图2

图3

图4

图5

图6

图7

图8

图9

图10

表1

图11

表2

图12

图13

图14

图15

表3

表4

图16

参考文献 25

1	Eggert D W, Fitzgibbon A W, Fisher R B. Simultaneous Registration of Multiple Range Views for Use in Reverse Engineering[C]//13th International Conference on Pattern Recognition. Vienna, Austria: IEEE, 1996: 243-247.
2	Wang Yuan, Yang Bisheng, Chen Yiping, et al. JoKDNet: A Joint Key Point Detection and Description Network for Large-scale Outdoor TLS Point Clouds Registration[J]. International Journal of Applied Earth Observation and Geoinformation (S0303-2434), 2021, 104: 102534.
3	Ren T, Wu R. An Acceleration Algorithm of 3D Point Cloud Registration Based on Iterative Closet Point[C]//2020 Asia-Pacific Conference on Image Processing, Electronics and Computers (IPEC). Dalian, China: IEEE, 2020: 271-276.
4	Zhao J, Chu J, Feng G, et al. Multimodal Point Cloud Registration Based on Adaptive Feature Region in Radiotherapy Guidance[C]//2020 International Conference on Virtual Reality and Visualization (ICVRV). Recife, Brazil: IEEE, 2020: 185-190.
5	Chang W, Zwicker M. Global Registration of Dynamic Range Scans for Articulated Model Reconstruction[J]. ACM Transactions on Graphics, 2011, 30(3): 26-40.
6	Yang Y, Ong S H, Foong K W C. A Robust Global and Local Mixture Distance Based Non-rigid Point Set Registration[J]. Pattern Recognition, 2015, 48(1): 156-173.
7	Salvi J, Matabosch C, Fofi D, et al. A Review of Recent Range Image Registration Methods with Accuracy Evaluation[J]. Image and Vision Computing, 2007, 25(5): 578-596.
8	Rusu R B, Blodow N, Beetz M. Fast Point Feature Histograms(FPFH) for 3D Registration[C]//2009 IEEE International Conference on Robotics and Automation. Kobe, Japan: IEEE, 2009: 3212-3217.
9	Aiger Dror, Mitra Niloy J, Cohen-Or Daniel. 4-points Congruent Sets for Robust Pairwise Surface Registration[J]. ACM Transactions on Graphics, 2008, 27(3): 1-10.
10	刘世光, 王海荣, 刘锦. 快速四点一致性点云粗配准算法[J]. 山东大学学报(工学版), 2019, 49(2):1-7.
	Liu Shiguang, Wang Hairong, Liu Jin. Fast Four-point Consistency Point Cloud Coarse Registration Algorithm[J]. Journal of Shandong University (Engineering Science), 2019, 49(2): 1-7.
11	彭真, 吕远健, 渠超, 等. 基于关键点提取与优化迭代最近点的点云配准[J]. 激光与光电子学进展, 2020, 57(6): 68-79.
	Peng Zhen, Yuanjian Lü, Qu Chao, et al. Point Cloud Registration Based on Key Point Extraction and Optimization Iteration[J]. Progress in Laser and Optoelectronics, 2020, 57(6): 68-79.
12	孙培芪, 卜俊洲, 陶庭叶, 等. 基于特征点法向量的点云配准算法[J]. 测绘通报, 2019(8): 48-53.
	Sun Peiqi, Bu Junzhou, Tao Tingye, et al. Point Cloud Registration Algorithm Based on Feature Point Normal Vector[J]. Bulletin of Surveying and Mapping, 2019(8): 48-53.
13	Lowe D G. Object Recognition from Local Scale-Invariant Features[C]//Seventh IEEE International Conference on Computer Vision. Kerkyra, Greece: IEEE, 1999: 1150-1157.
14	Lowe D G. Distinctive Image Features from Scale-invariant Key Points[J]. International Journal of Computer Vision(S0920-5691), 2004, 60(2): 91-110.
15	Besl P J, McKay N D. A Method for Registration of 3-D Shapes[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence (S1939-3539), 1992, 14(2): 239-256.
16	孙水发, 李准, 夏坤, 等. 变尺度点云配准算法[J]. 系统仿真学报, 2018, 30(7): 2465-2474.
	Sun Shuifa, Li Zhun, Xia Kun, et al. Variable-scale Point Cloud Registration Algorithm[J]. Journal of System Simulation, 2018, 30(7): 2465-2474.
17	石珣, 任洁, 任小康, 等. 基于曲率特征的漂移配准方法[J]. 激光与光电子学进展, 2018, 55(8): 248-254.
	Shi Xun, Ren Jie, Ren Xiaokang, et al. Drift Registration Method Based on Curvature Characteristics[J]. Advances in Laser and Optoelectronics, 2018, 55(8): 248-254.
18	王飞鹏, 肖俊, 王颖, 等. 一种基于高斯曲率的ICP改进算法[J]. 中国科学院大学学报, 2019, 36(5): 702-708.
	Wang Feipeng, Xiao Jun, Wang Ying, et al. An Improved ICP Algorithm Based on Gaussian Curvature[J]. Journal of University of Chinese Academy of Sciences, 2019, 36(5): 702-708.
19	Yao Z, Zhao Q, Li X, et al. Point Cloud Registration Algorithm Based on Curvature Feature Similarity[J]. Measurement(S0263-2241), 2021, 177(11): 109274.
20	Hu F J, Zhao Y W, Chen J. SIFT Feature Points Detection and Extraction of Three-dimensional Point Cloud[C]//International Conference on Communication Technology and Application. Beijing: WIT Press, 2014: 603-611.
21	Chung D H, Yun I D, Sang U L, et al. Registration of Multiple-range Views Using the Reverse-calibration Technique[J]. Pattern Recognition, 1998, 31(4): 457-464.
22	Sorkine-Hornung O, Rabinovich M. Least-squares Rigid Motion Using Svd[C]//International Conference on Computational Science. Computing, 2017: 1-5.
23	Wu Z, Song S, Khosla A, et al. 3D ShapeNets: A Deep Representation for Volumetric Shapes[C]//2015 IEEE Conference on Computer Vision and Pattern Recognition. Boston, MA, USA: IEEE, 2015: 1912-1920.
24	Xu G, Pang Y, Bai Z, et al. A Fast Point Clouds Registration Algorithm for Laser Scanners[J]. Applied Sciences(S2076-3417), 2021, 11(8): 3426.
25	梁振斌, 熊风光, 韩燮, 等. 基于深度学习的点云匹配[J]. 计算机工程与设计, 2020, 41(6): 1689-1693.
	Liang Zhenbin, Xiong Fengguang, Han Xie, et al. Point-Cloud Matching Based on Deep Learning[J]. Computer Engineering and Design, 2020, 41(6): 1689-1693.

模型	算法	牛	龙	茶壶
原始	3D-SIFT	4.350	3.540	3.381
原始	Harris	12.186	11.564	23.836
旋转平移	3D-SIFT	13.419	10.391	13.013
旋转平移	Harris	15.416	15.970	20.287
尺度缩小	3D-SIFT	8.973	9.185	7.122
尺度缩小	Harris	13.111	12.044	48.784

模型	旋转平移	尺寸缩小
模型	改进前/改进后	改进前/改进后
鸟	83.6/88.5	100/100
牛	74.3/84.0	91.9/92.8
龙	29.0/66.4	70.1/89.5

算法	鸟	牛	龙
FPFH	281.028	57.525	406.530
ICP^[15]	19.764	2.024	9.404
改进的ICP算法^[24]	1.608	1.131	1.390
本文	62.451	22.741	116.958

模型	匹配方法	时间/s	旋转误差	平移误差
帽子	ICP^[15]	1.125	19.4	0.012 7
	基于Marvin的点云匹配^[25]	0.761	8.7	0.002 9
	本文算法	11.290	11.9	0.000 7
椅子	ICP^[15]	1.279	16.2	0.023 7
	基于Marvin的点云匹配^[25]	0.828	7.0	0.002 4
	本文算法	13.225	11.8	0.000 3

[1]	上官晋太, 党雅文, 连玮. 非刚体图像配准中一种改进的点匹配方法[J]. 系统仿真学报, 2022, 34(7): 1482-1489.
[2]	魏亮, 薛牧遥, 霍炬, 张金杰. 基于ICP算法的非合作目标特征点云配准优化[J]. 系统仿真学报, 2020, 32(12): 2383-2387.
[3]	顾菊平, 程天宇, 王建平, 华亮, 赵凤申, 蒋凌. 基于几何特征不变量的快速3D医学图像配准[J]. 系统仿真学报, 2020, 32(11): 2105-2111.

基于关键点精确配对的点云曲面匹配方法

Point Cloud Surface Matching Method Based on Precise Matching of Critical Point

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 20

参考文献 25

相关文章 3

编辑推荐

Metrics

本文评价