[1]姬硕,胡立华,张素兰,等.基于双重注意力和匹配矩阵优化的点云配准算法[J].计算机技术与发展,2025,(05):97-105.[doi:10.20165/j.cnki.ISSN1673-629X.2025.0002]
　JI Shuo,HU Li-hua,ZHANG Su-lan,et al.Point Cloud Registration Algorithm Based on Dual Attention and Matching Matrix Optimization[J].,2025,(05):97-105.[doi:10.20165/j.cnki.ISSN1673-629X.2025.0002]
点击复制

基于双重注意力和匹配矩阵优化的点云配准算法()

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2025年05期
页码:
97-105
栏目:
人工智能
出版日期:
2025-05-10

文章信息/Info

Title:
Point Cloud Registration Algorithm Based on Dual Attention and Matching Matrix Optimization
文章编号:
1673-629X(2025)05-0097-09
作者:
姬硕1; 胡立华1; 张素兰1; 胡建华2; 3; 王欣波2; 3
1. 太原科技大学 计算机科学与技术学院,山西 太原 030024;
2. 中国科学院 自动化研究所,北京 100190;
3. 中科锐智(洛阳)数码科技有限公司,河南 洛阳 471000
Author(s):
JI Shuo1; HU Li-hua1; ZHANG Su-lan1; HU Jian-hua2; 3; WANG Xin-bo2; 3
1. School of Computer Science and Technology,Taiyuan University of Science and Technology,Taiyuan 030024,China;
2. Institute of Automation,Chinese Academy of Sciences,Beijing 100190,China;
3. Zhongke Ruizhi (Luoyang) Digital Technology Co. ,Ltd. ,Luoyang 471000,China
关键词:
点云配准; 通道注意力; 空间注意力; 匹配矩阵优化; 深度学习
Keywords:
point cloud registration; channel attention; spatial attention; matching matrix optimization; deep learning
分类号:
TP391
DOI:
10.20165/j.cnki.ISSN1673-629X.2025.0002
摘要:
针对点云配准过程中由于噪声、误匹配和漏匹配导致点云配准算法的配准精度低、鲁棒性差的问题,提出了一种融合双重注意力和匹配矩阵优化的点云配准算法。 首先,设计了结合通道注意力和空间注意力的双重注意力模块,对噪声部分赋予较低的权重,使模型能够更专注于重要或相关的信息,从而减少噪声对模型的影响。 其次,结合特征点的局部信息和全局信息设计了匹配矩阵优化模块,模型可以充分利用点云数据的多层次特征,从而提高配准的准确性。 最后,以人工合成数据集 ModelNet40、真实室内场景数据集7Scenes 和真实室外场景数据集 KITTI 为对象进行验证,在 ModelNet40 高噪声、7Scenes 和 KITTI 的点云配准实验中,旋转矩阵和平移向量的均方根误差分别降低至 0. 665 7 和 0. 001 7、0. 079 6 和 0. 000 9、2. 061 7 和 0. 041 7。 实验结果表明,该方法可以在降低噪声对模型影响的同时,有效地减少漏匹配以及剔除误匹配,提高点云配准的精度和鲁棒性。
Abstract:
To address the issues of low registration accuracy and poor robustness in point cloud registration algorithms caused by noise,mismatches,and missing correspondences,a point cloud registration algorithm integrating dual attention and matching matrix optimization is proposed. Firstly,to improve the registration results caused by noise points,a dual attention module combining channel attention and spatial attention is designed,where noisy points are assigned lower weights,this allows the model to focus more on important or relevant information. Secondly,a matching matrix optimization module is designed by incorporating both local and global information of feature points,enabling the model to fully utilize the multi-level features of the point cloud data, thereby improving the registration accuracy.Finally,validation is conducted on the synthetic dataset ModelNet40,the real-world indoor dataset 7Scenes,and the real-world outdoor dataset KITTI. In the point cloud registration experiments on ModelNet40 with high noise,7Scenes,and KITTI,the root mean square errors of the rotation matrix and translation vector were reduced to 0. 665 7 and 0. 001 7,0. 079 6 and 0. 000 9,2. 061 7 and 0. 041 7,re-spectively. The experimental results demonstrate that the proposed method effectively reduces missed matches and eliminates mismatches while minimizing the influence of noise on the model,thereby improving the accuracy and robustness of point cloud registration.

相似文献/References:

[1]谢晓燕,吴锦桥. 一种全自动三维点云配准及比例约束方法[J].计算机技术与发展,2015,25(03):63.
　XIE Xiao-yan,WU Jin-qiao. An Automatic Method of 3 D Point Cloud Registration and Dimension Adjustment[J].,2015,25(05):63.
[2]张 伟,刘宁钟,寇金桥.基于深度特征金字塔的路面病害检测[J].计算机技术与发展,2022,32(12):173.[doi:10. 3969 / j. issn. 1673-629X. 2022. 12. 026]
　ZHANG Wei,LIU Ning-zhong,KOU Jin-qiao.Pavement Disease Detection Based on Depth Feature Pyramids[J].,2022,32(05):173.[doi:10. 3969 / j. issn. 1673-629X. 2022. 12. 026]
[3]谢紫薇,鲁大营*,李志琦,等.基于扩张卷积与注意力的甲状腺超声分割方法[J].计算机技术与发展,2023,33(03):71.[doi:10. 3969 / j. issn. 1673-629X. 2023. 03. 011]
　XIE Zi-wei,LU Da-ying*,LI Zhi-qi,et al.Dilated Convolution and Attention-based Ultrasound Segmentation of Thyroid Nodules[J].,2023,33(05):71.[doi:10. 3969 / j. issn. 1673-629X. 2023. 03. 011]
[4]关 慧,曹同洲.基于 CNN 和多注意力机制的 XSS 检测模型[J].计算机技术与发展,2023,33(04):175.[doi:10. 3969 / j. issn. 1673-629X. 2023. 04. 026]
　GUAN Hui,CAO Tong-zhou.XSS Detection Model Based on CNN and Multi-attention Mechanism[J].,2023,33(05):175.[doi:10. 3969 / j. issn. 1673-629X. 2023. 04. 026]
[5]马 卫*,李微微.改进人工蜂群算法的点云数据配准优化研究[J].计算机技术与发展,2023,33(06):79.[doi:10. 3969 / j. issn. 1673-629X. 2023. 06. 013]
　MA Wei*,LI Wei-wei.Research on Point Clouds Data Registration Optimization Based on Improved Artificial Bee Colony Algorithm[J].,2023,33(05):79.[doi:10. 3969 / j. issn. 1673-629X. 2023. 06. 013]
[6]周 帅,李 理,彭章君,等.基于多通道特征和混合注意力的环境声音分类[J].计算机技术与发展,2023,33(08):43.[doi:10. 3969 / j. issn. 1673-629X. 2023. 08. 007]
　ZHOU Shuai,LI Li,PENG Zhang-jun,et al.Environmental Sound Classification Based on Multi-channel Features and Mixed Attention[J].,2023,33(05):43.[doi:10. 3969 / j. issn. 1673-629X. 2023. 08. 007]
[7]杨 虹,范 勇.一种基于区分区域定位的细粒度图像识别方法[J].计算机技术与发展,2023,33(11):169.[doi:10. 3969 / j. issn. 1673-629X. 2023. 11. 025]
　YANG Hong,FAN Yong.A Fine Grain Image Recognition Method Based on Distinguishable Region Location[J].,2023,33(05):169.[doi:10. 3969 / j. issn. 1673-629X. 2023. 11. 025]
[8]宁园园,张素兰,陈 飞.基于双注意力机制的零样本建筑图像分类方法[J].计算机技术与发展,2023,33(10):35.[doi:10. 3969 / j. issn. 1673-629X. 2023. 10. 006]
　NING Yuan-yuan,ZHANG Su-lan,CHEN Fei.Zero-shot Architectural Image Classification Method Based on Dual Attention Mechanism[J].,2023,33(05):35.[doi:10. 3969 / j. issn. 1673-629X. 2023. 10. 006]
[9]姚亮亮,张太红*,张洋宁,等.单参数通道注意力模块[J].计算机技术与发展,2023,33(12):215.[doi:10. 3969 / j. issn. 1673-629X. 2023. 12. 030]
　YAO Liang-liang,ZHANG Tai-hong*,ZHANG Yang-ning,et al.Single-parameter Channel Attention Module[J].,2023,33(05):215.[doi:10. 3969 / j. issn. 1673-629X. 2023. 12. 030]
[10]尹春勇,沈子宁.基于交互式特征与多尺度特征的文本相似度研究[J].计算机技术与发展,2024,34(08):86.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0140]
　YIN Chun-yong,SHEN Zi-ning.Research on Text Similarity Based on Interactive Features and Multi-scale Features[J].,2024,34(05):86.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0140]

更新日期/Last Update: 2025-05-10