环视BEV多层级语义车位自适应加权配准算法

史兴领; 朱添翼; 黄刘生; 檀杰; 赵晓东

doi:null

您当前的位置：

首页 >

文章列表页 >

环视BEV多层级语义车位自适应加权配准算法

光学传感与测量 | 更新时间：2026-01-22

- 环视BEV多层级语义车位自适应加权配准算法
- Multi-level semantic parking spot adaptive weighted registration algorithm for Surround-view BEV
- 光学技术 2026年52卷第1期页码：109-114
- 作者机构：
  
  1.中国科学技术大学计算机科学与技术学院，安徽合肥 230027
  2.科大国创合肥智能汽车科技有限公司感知融合部，安徽合肥 230088
- 作者简介：
  
  史兴领（1971— ），男，高级工程师，硕士，从事人工智能、自动控制、软件工程与算法以及智能网联汽车相关技术和产品的研发与项目管理工作。xlshi@ustcsoft.com
  朱添翼（1992— ），男，中级工程师，硕士，从事摄影测量与遥感和多传感器融合应用研究。zhu.ty@qq.com
- 基金信息：
  
  安徽省科技攻关计划项目(202423k09020037)
- DOI：
  中图分类号： TP751
- 收稿：2025-05-29，
  
  修回：2025-12-07，
  
  纸质出版：2026-01-20
- 稿件说明：
移动端阅览
史兴领,朱添翼,黄刘生等.环视BEV多层级语义车位自适应加权配准算法[J].光学技术,2026,52(01):109-114.

SHI Xinling,ZHU Tianyi,HUANG Liusheng,et al.Multi-level semantic parking spot adaptive weighted registration algorithm for Surround-view BEV[J].OPTICAL TECHNIQUE,2026,52(01):109-114.
史兴领,朱添翼,黄刘生等.环视BEV多层级语义车位自适应加权配准算法[J].光学技术,2026,52(01):109-114. DOI：

SHI Xinling,ZHU Tianyi,HUANG Liusheng,et al.Multi-level semantic parking spot adaptive weighted registration algorithm for Surround-view BEV[J].OPTICAL TECHNIQUE,2026,52(01):109-114. DOI：

摘要

车载BEV空间配准技术作为智能驾驶车辆的重要手段，一直是自动驾驶领域研究的热点。在BEV空间，如何自适应的提高配准精度，成为自动驾驶BEV空间应用的难点。针对视觉数据深度估计误差会随距离累积、激光数据在低强度场景鲁棒性差和多源配准强依赖于时间同步精度，时序偏差易引发错位等问题，提出环视BEV多层级语义车位加权配准方法，即基于环视BEV拼接图后的语义识别结果，获取车位等语义特征信息，对车位进行多层级跟踪，确定配准初始位姿，基于车载传感器成像精度与距离相关原理，配准平差中引入基于欧式距离的自适应权重计算，从而提高配准精度和鲁棒性。该方法极大提高了配准精度，鲁棒性高，能够较好的基于配准结果进行建图与定位。地下车库环视数据集的建图与定位结果验证了利用本算法进行BEV空间配准的精度与效果。

Abstract

In-vehicle BEV （Bird's Eye View） spatial registration technology is an important tool for intelligent driving vehicles and has consistently been a research focus in the field of autonomous driving. In the BEV space

adaptively improving registration accuracy has become a major challenge for BEV applications in autonomous driving. Issues such as the accumulation of depth estimation errors in visual data over distance

the poor robustness of LiDAR data in low-intensity scenarios

and the strong reliance of multi-source registration on time synchronization

where temporal deviations easily lead to misalignment are addressed. A surround-view BEV multi-level semantic parking space weighted registration method is proposed. Specifically

based on the semantic recognition results from the surround-view BEV stitched image

semantic feature information such as parking spaces is obtained

multi-level tracking of these parking spaces is carried out to determine the initial registration pose

and adaptive weights based on Euclidean distance are introduced into the registration adjustment according to the imaging accuracy of in-vehicle sensors and distance-related principles. This enhances registration accuracy and robustness. The method significantly improves registration precision and robustness

enabling effective mapping and localization based on the registration results. The mapping and localization results from an underground parking garage surround-view dataset validate the accuracy and effectiveness of BEV spatial registration using this algorithm.

关键词

Keywords

references

Clavijo M , Jiménez Felipe , et al . The development and prospects of autonomous driving technology [J ] . Applied Sciences , 2023 , 13 （ 9 ）: 2076 — 3417 .

Yang B , Li J , Zeng T . A review of environmental perception technology based on multi-sensor information fusion in autonomous driving [J ] . World Electric Vehicle Journal , 2025 , 16 （ 1 ）: 20 .

林俊钦 , 胡晓明 , 闫达远 . 场景轮廓的动态规划立体匹配算法 [J ] . 光学技术 , 2013 （ 1 ）: 5 .

Lin Junqin , Hu Xiaoming , Yan Dayuan . Dynamic programming stereo matching algorithm for scene profiles [J ] . Optical Technology , 2013 （ 1 ）: 5 .

Zhao J , Shi J , Zhuo L . BEV perception for autonomous driving: State of the art and future perspectives [J ] . Expert Systems with Applications , 2024 , 258 （ 000 ）: 18 .

Wei M , Li J , Kang H , et al . BEV-CFKT: A LiDAR-camera cross-modality-interaction fusion and knowledge transfer framework with transformer for BEV 3D object detection [J ] . Neurocomputing , 2024 , 582 : 127527 .

赵园园 , 曾亚涛 . 自动驾驶下的纯视觉BEV融合研究与应用 [J ] . 计算机应用文摘 , 2024 , 40 （ 14 ）: 107 — 108 .

Zhao Yuanyuan , Zeng Yatao . Research and application of pure visual BEV fusion under autonomous driving [J ] . Computer Application Abstracts , 2024 , 40 （ 14 ）: 107 — 108 .

Samani E U , Tao F , Dasari H R , et al . F2BEV: bird's eye view generation from surround-view fisheye camera images for automated driving [C ] // 2023 IEEE/RSJ International Conference on Intelligent Robots and Systems （IROS） . Detroit, MI, USA : IEEE , 2023 : 9367 — 9374 .

刘俊杰 . 跨模态引导的纯视觉3D感知算法的研究 [D ] . 广州 : 华南理工大学 , 2023 .

Liu Junjie . Research on pure visual 3D perception algorithms based on cross-modal guidance [D ] . Guangzhou : South China University of Technology , 2023 .

刘天弋 . 面向城市道路环境的车载低线数LiDAR/INS/GNSS融合导航算法研究 [D ] . 武汉：武汉大学 , 2023 .

Liu Tianyi . Research on the fusion navigation algorithm of low-line LiDAR/INS/GNSS in vehicle-mounted low-lines for urban road environments [D ] . Wuhan : Wuhan University , 2023 .

Park G , Koh J , Kim J , et al . LiDAR-based 3D temporal object detection via Motion-aware LiDAR feature fusion [J ] . Sensors （ 14248220 ）, 2024 , 24 （ 14 ）: 4667 .

Kim M , Kim G , Jin K H , et al . BroadBEV: collaborative LiDAR-camera fusion for Broad-sighted bird's eye view map construction [C ] // 2023 IEEE International Conference on Robotics and Automation （ICRA 2023） . London,UK : IEEE , 2023 : 11125 — 11132 .

Cai H . BEVFusion 4 D: learning LiDAR-camera fusion under bird's-Eye-View via cross-modality guidance and temporal aggregation[EB/OL ] . （ 2023-03 ）/[ 2025-05 ] . https://api.semanticscholar.org/CorpusID:257833799 https://api.semanticscholar.org/CorpusID:257833799 .

Ronneberger O , Fischer P , Brox T . U- Net: Convolutional networks for biomedical image segmentation [C ] // Medical Image Computing and Computer-Assisted Intervention-MICCAI 2015 . Munich, Germany . Cham : Springer International Publishing , 2015 : 237 — 244 .

杨颖 , 岳圣斌 , 楚博文 , 等 . 同尺度和跨尺度增强的U-Net用于视网膜血管分割 [J ] . 光学技术 , 2023 , 49 （ 4 ）: 487 — 496 .

Yang Ying , Yue Shengbin , Chu Bowen , et al . U-Net with the same scale and cross-scale enhancement is used for retinal vascular segmentation [J ] . Optical Technology , 2023 , 49 （ 4 ）: 487 — 496 .

刘利源 . 基于多层级语义信息适配的图像语义分割 [D ] . 天津 : 天津大学 , 2020 .

Liu Liyuan . Image semantic segmentation based on multi-level semantic information adaptation [D ] . Tianjin : Tianjin University , 2020 .

高智 , 胡傲涵 , 陈泊安 , 等 . 多层级几何—语义融合的图神经网络地表异常检测框架 [J ] . 通感学报 , 2024 , 28 （ 7 ）: 2051 — 2068 .

Gao Zhi , Hu Aohan , Chen Bo'an , et al . Multi-level geometric- semantic fusion graph neural network surface anomaly detection framework [J ] . Journal of Remote Sensing , 2024 , 28 （ 7 ）: 2051 — 2068 .

占钟鸣 , 李庆武 , 余大兵 , 等 . 基于边缘及多尺度特征融合的显著性目标检测方法 [J ] . 光学技术 , 2024 , 50 （ 5 ）: 606 — 612 .

Zhan Zhongming , Li Qingwu , Yu Dabing , et al . Senseful target etection method based on edge and multi-scale feature fusion [J ] . Optical Technology , 2024 , 50 （ 5 ）: 606 — 612 .

Huiyan W , Tengfei Z , Fumin M A . Rough K-means algorithm with self-adaptive weights measurement based on space distance [J ] . Computer Science , 2018 , 45 （ 7 ）: 190 — 196 .

周辉奎 , 顾牡丹 . 自适应加权直方图均衡化的红外图像增强 [J ] . 光学技术 , 2023 , 49 （ 6 ）: 750 — 755 .

Zhou Huikui , Gu Mudan . Infrared image enhancement of adaptive weighted histogram equalization [J ] . Optical Technology , 2023 , 49 （ 6 ）: 750 — 755 .

Dereniak E L , Dereniak T D . Geometrical and trigonometric optics: image formation [M ] . New York : Cambridge University Press , 2008 .

Jiang S , Qi P , Han L , et al . Navigation system for orchard spraying robot based on 3D LiDAR SLAM with NDT_ICP point cloud registration [J ] . Computers and Electronics in Agriculture , 2024 : 220 : 108870 .

王海起 , 张腾 , 彭佳琦 , 等 . 空间加权距离的GIS数据Fuzzy C—means聚类方法与应用分析 [J ] . 地球信息科学学报 , 2013 , 15 （ 4 ）: 536 — 542 .

Wang Haiqi , Zhang Teng , Peng Jiaqi , et al . Fuzzy C-means clustering method and application analysis of GIS data with space-weighted distances [J ] . Journal of Earth Information Science , 2013 , 15 （ 4 ）: 536 — 542 .

Liu Z , Zhu Y , Zhang C . Change detection method of high resolution remote sensing image based on unet network [C ] // 2024 5th International Conference on Intelligent Computing and Human-Computer Interaction （ICHCI） . Nanchang,China （Virtual） : IEEE , 2024 : 151 — 157 .

Zhao X , Yuan Y , Song M , et al . Use of unmanned aerial vehicle imagery and deep learning unet to extract rice lodging [J ] . Sensors （ 14248220 ）, 2019 , 19 （ 18 ）: 3859 .

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

暂无数据