面向智能驾驶的自适应多尺度三维点云配准网络

汪霖; 韩栋; 韩景怡; 陈圣梅; 刘成

doi:null

您当前的位置：

首页 >

文章列表页 >

面向智能驾驶的自适应多尺度三维点云配准网络

更新时间：2026-06-19

- 面向智能驾驶的自适应多尺度三维点云配准网络
- Adaptive Multi-scale 3D Point Cloud Registration Network for Autonomous Driving
- 西安交通大学学报 2026年
- 作者机构：
  
  西北大学电子信息学院，710127，西安
- 作者简介：
- 基金信息：
- DOI：
  中图分类号： TP29
- 收稿：2026-04-21，
  
  修回：2026-06-17，
  
  录用：2026-06-19，
- 稿件说明：
移动端阅览
汪霖, 韩栋, 韩景怡, 等. 面向智能驾驶的自适应多尺度三维点云配准网络[J/OL]. 西安交通大学学报, 2026.

WANG Lin, HAN Dong, HAN Jinyi, et al. Adaptive Multi-scale 3D Point Cloud Registration Network for Autonomous Driving[J/OL]. JOURNAL OF XI’AN JIAOTONG UNIVERSITY, 2026.
汪霖, 韩栋, 韩景怡, 等. 面向智能驾驶的自适应多尺度三维点云配准网络[J/OL]. 西安交通大学学报, 2026. DOI：

WANG Lin, HAN Dong, HAN Jinyi, et al. Adaptive Multi-scale 3D Point Cloud Registration Network for Autonomous Driving[J/OL]. JOURNAL OF XI’AN JIAOTONG UNIVERSITY, 2026. DOI：

摘要

针对智能驾驶场景中激光雷达点云稀疏、特征模糊以及低重叠引发关键点误匹配的问题，提出了一种面向智能驾驶的自适应多尺度三维点云配准网络。首先，该网络利用自适应多尺度特征提取模块获取采样点的多层级特征，并结合自适应邻域稳定器实现关键点的精确定位，从而解决由于特征模糊而导致的匹配难题。其次，通过融合多尺度特征匹配一致性与几何一致性校验，提升待配准点对的内点置信度。最后，基于该置信度，采用由粗到精的迭代优化策略，完成三维点云的精确配准。KITTI数据集上的测试结果表明：该网络的相对平移误差和相对旋转误差分别低至0.0618 m和0.21°，配准召回率高达99.8%。nuScenes数据集上的测试结果表明：在帧间距离分别为10、20、30 m下，配准召回率分别达到90.1%、83.3%和78.8%。上述结果验证了该网络能够有效抑制复杂道路场景下的误匹配，实现高精度、高鲁棒性的大规模三维点云配准。该研究适配了动态复杂路况下的实际应用需求，为自动驾驶场景下的精准定位、环境感知提供可靠的技术支撑。

Abstract

To address the problems of sparse LiDAR point clouds

ambiguous features

and mismatches of keypoints caused by low overlap in intelligent driving scenarios

this paper proposes an adaptive multi-scale 3D point cloud registration network for intelligent driving. First

the network employs an adaptive multi-scale feature extraction module to obtain multi-level features of sampled points

combined with an adaptive neighborhood stabilizer for precise keypoint localization

thereby resolving matching difficulties caused by feature ambiguity. Second

by integrating multi-scale feature matching consistency with geometric consistency verification

the inlier confidence of point pairs to be registered is improved. Finally

based on this confidence

a coarse-to-fine iterative optimization strategy is adopted to achieve accurate 3D point cloud registration. Test results on the KITTI dataset demonstrate that the network achieves a relative translation error as low as 0.0618 m and a relative rotation error as low as 0.21°

with a registration recall of up to 99.8%. Results on the nuScenes dataset indicate that with inter-frame distances of 10 m

20 m

and 30 m

the registration recall reaches 90.1%

83.3%

and 78.8%

respectively. These findings validate that the proposed network effectively suppresses mismatches in complex road scenarios

achieving high-precision and highly robust large-scale 3D point cloud registration. This research meets the practical application requirements of dynamic and complex traffic conditions

providing reliable technical support for precise localization and environmental perception in autonomous driving scenarios.

关键词

Keywords

references

YU Z , ZHANG P , SHI J . Transformation of industrial robotics with natural language models: Recent progress and future prospects [J ] . Robotics and Computer Integrated Manufacturing , 2026 , 97 : 103113 - 103151 .

周翔 , 宋林烨 , 夏海峰 , 等 . 个性化热舒适系统在电动汽车中的应用与展望 [J ] . 西安交通大学学报 , 2026 , 60 ( 1 ): 1 - 13 .

ZHOU X , SONG L , XIA H , et al . Application and prospect of personalized thermal comfort system in electric vehicles [J ] . Journal of Xi'an Jiaotong University , 2026 , 60 ( 1 ): 1 - 13 .

SONG C H , BLUKIS V , TREMBLAY J , et al . Robospatial: Teaching spatial understanding to 2D and 3D vision-language models for robotics [C ] // Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2025 : 15768 - 15780 .

WANG Z , SU Y , LI C , et al . Open-vocabulary octree-graph for 3D scene understanding [C ] // Proceedings of the IEEE/CVF International Conference on Computer Vision , 2025 : 7037 - 7047 .

马庆禄 , 白锋 , 张杰 , 等 . 基于激光雷达点云地图的车辆定位与导航 [J ] . 光学精密工程 , 2024 , 32 ( 16 ): 2537 - 2549 .

MA Q , BAI F , ZHANG J , et al . Vehicle localization and navigation based on LiDAR point cloud map [J ] . Optics and Precision Engineering , 2024 , 32 ( 16 ): 2537 - 2549 .

MAKEN F A , KARMAKA M , PRABHU V , et al . Improving 3D reconstruction through RGB-D sensor uncertainty estimation and noise modeling [J ] . Sensors , 2025 , 25 ( 3 ): 950 - 966 .

TIAN Y , LI X , YIN J . From pseudo-to non-correspondences: robust point cloud registration via thickness-guided self-correction [J ] . Computers & Graphics , 2026 : 104521 - 104540 .

MA K , YAN F , LI S , et al . Low-overlap registration of multi-source LiDAR point clouds in urban scenes through dual-stage feature pruning and progressive hierarchical methods [J ] . Remote Sensing , 2025 , 17 ( 17 ): 2938 - 2955 .

KASHEF TABRIZIAN S , TERRYN S , VANDERBORGHT B . Toward autonomous self-healing in soft robotics: a review and perspective for future research [J ] . Advanced Intelligent Systems , 2025 , 7 ( 8 ): 2400790 - 2400815 .

王丞 , 田暄 , 郭瑞 , 等 . 自适应Harris角点提取的点云粗配准算法 [J ] . 西安交通大学学报 , 2022 , 56 ( 3 ): 33 - 44 .

WANG C , TIAN X , GUO R , et al . A coarse registration algorithm for point clouds based on adaptive Harris corner detection [J ] . Journal of Xi'an Jiaotong University , 2022 , 56 ( 3 ): 33 - 44 .

WANG W , MEI G , ZHANG J , et al . Fully-geometric cross-attention for point cloud registration [C ] // 2025 International Conference on 3D Vision (3DV) , IEEE , 2025 : 347 - 356 .

YAQI X , YAN X , et al . ASFM-Net: Asymmetrical siamese feature matching network for point completion [C ] // Proceedings of the 29th ACM International Conference on Multimedia , 2021 : 1938 - 1947 .

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 .

WANG Y , SOLOMON J M . Deep closest point: Learning representations for point cloud registration [C ] // Proceedings of the IEEE/CVF International Conference on Computer Vision , 2019 : 3523 - 3532 .

YEW Z J , LEE G H . RPM-Net: Robust point matching using learned features [C ] // Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2020 : 11824 - 11833 .

FU K , LIU S , LUO X , et al . Robust point cloud registration framework based on deep graph matching [C ] // Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2021 : 8893 - 8902 .

BAI X , LUO Z , ZHOU L , et al . PointDSC: Robust point cloud registration using deep spatial consistency [C ] // Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2021 : 15859 - 15869 .

YEW Z J , LEE G H . REGTR: end-to-end point cloud correspondences with transformers [C ] // Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2022 : 6677 - 6686 .

QIN Z , YU H , WANG C , et al . Geometric transformer for fast and robust point cloud registration [C ] // Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2022 : 11143 - 11152 .

QI C R , YI L , SU H , et al . PointNet++: deep hierarchical feature learning on point sets in a metric space [J ] . Advances in Neural Information Processing Systems , 2017 : 30 - 39 .

GEIGER A , LENZ P , URTASUN R . Are we ready for autonomous driving? the KITTI vision benchmark suite [C ] // 2012 IEEE Conference on Computer Vision and Pattern Recognition , 2012 : 3354 - 3361 .

YU H , LI F , SALEH M , et al . CoFiNet: reliable coarse-to-fine correspondences for robust point cloud registration [J ] . Advances in Neural Information Processing Systems , 2021 , 34 : 23872 - 23884 .

HUANG S , GOJCIC Z , USVYATSOV M , et al . PREDATOR: registration of 3D point clouds with low overlap [C ] // Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2021 : 4267 - 4276 .

YAN S , SHI P , ZHAO Z , et al . TurboReg: turboclique for robust and efficient point cloud registration [C ] // Proceedings of the IEEE/CVF International Conference on Computer Vision , 2025 : 26371 - 26381 .

CAESAR H , BANKITI V , et al . NuScenes: a multimodal dataset for autonomous driving [C ] // Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2020 : 11621 - 11631 .

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

一种可演化网络的研究与实践

自适应移频变尺度随机共振在故障诊断中的应用

判断运动复杂度的快速运动估计算法

毫米波通信系统中利用随机逼近的波束赋形算法

自适应折叠混沌优化方法