通信延迟下车辆协同感知的3D目标检测方法

doi:10.3778/j.issn.1002-8331.2311-0280

摘要/Abstract

摘要： 针对车辆协同感知3D目标检测在通信延迟条件下精度较低的问题，提出一种通信延迟下车辆协同感知的3D目标检测方法。设计时空预测模块，提取通信延迟车辆历史感知特征序列中的时空特征，以预测当前时刻的感知特征，基于预测特征构建感知融合模块，利用注意力机制动态融合感知特征，以降低预测误差影响，提高检测精度。该方法在OPV2V、V2XSet和V2V4Real数据集上进行实验，与Where2Comm、V2VNet等主流协同感知方法相比。实验结果表明，在所对比的方法中，Where2Comm在不同延迟下的3D目标检测平均精度最优，该方法相比Where2Comm在400 ms下的平均精度分别提高了5.9、3.9和1.5个百分点。

关键词: 协同感知, 3D目标检测, 通信延迟, 时空序列预测, 注意力机制, 特征融合

Abstract: Aiming at the problem of low accuracy of 3D target detection for cooperative vehicle perception under communication delay conditions, a 3D target detection method for cooperative vehicle perception under communication delay is proposed. Firstly, a spatio-temporal prediction module is designed to extract the spatio-temporal features in the sequence of historical perception features of communication delayed vehicles in order to predict the perception features at the current moment, and then a perception fusion module is constructed based on the predicted features, which dynamically fuses the perception features using the attention mechanism to reduce the impact of prediction error and improve the detection accuracy. The method in this paper is experimented on OPV2V, V2XSet and V2V4Real datasets, compared with mainstream cooperative perception methods such as Where2Comm and V2VNet. The experimental results show that Where2Comm has the best average accuracy for 3D target detection at different delays among the compared methods, and the method improves the average accuracy by 5.9, 3.9 and 1.5 percentage points compared to Where2Comm at 400 ms, respectively.

Key words: collaborative perception, 3D object detection, communication delay, spatio-temporal series prediction, attention mechanism, feature fusion

卢敏, 胡振宇. 通信延迟下车辆协同感知的3D目标检测方法[J]. 计算机工程与应用, 2025, 61(7): 278-287.

LU Min, HU Zhenyu. 3D Object Detection Method for Cooperative Vehicle Sensing Under Communication Delay[J]. Computer Engineering and Applications, 2025, 61(7): 278-287.

参考文献

[1] DROBNITZKY M, FRIEDERICH J, EGGER B, et al. Survey and systematization of 3D object detection models and methods[J]. The Visual Computer, 2024, 40(3): 1867-1913.
[2] 崔靖. 基于深度学习的自动驾驶车辆三维目标检测方法研究[D]. 太原: 中北大学, 2022.
CUI J. Research on 3D object detection method for autonomous vehicle based on deep learning[D]. Taiyuan: North University of China, 2022.
[3] HAN Y S, ZHANG H, LI H F, et al. Collaborative perception in autonomous driving: methods, datasets and challenges[J]. arXiv:2301.06262, 2023.
[4] 杜昊. 面向车队自动驾驶的协同感知与避障策略研究[D]. 成都: 电子科技大学, 2023.
DU H. Research on collaborative sensing and obstacle avoidance strategy for platoon based autonomous driving[D]. Chengdu: University of Electronic Science and Technology of China, 2023.
[5] ARNOLD E, DIANATI M, DE TEMPLE R, et al. Cooperative perception for 3D object detection in driving scenarios using infrastructure sensors[J]. IEEE Transactions on Intelligent Transportation Systems, 2022, 23(3): 1852-1864.
[6] THANDAVARAYAN G, SEPULCRE M, GOZALVEZ J. Generation of cooperative perception messages for connected and automated vehicles[J]. IEEE Transactions on Vehicular Technology, 2020, 69(12): 16336-16341.
[7] CHEN Q, MA X, TANG S H, et al. F-cooper: feature based cooperative perception for autonomous vehicle edge computing system using 3D point clouds[C]//Proceedings of the 4th ACM/IEEE Symposium on Edge Computing. New York: ACM, 2019: 88-100.
[8] CHEN Q, TANG S H, YANG Q, et al. Cooper: cooperative perception for connected autonomous vehicles based on 3D point clouds[C]//Proceedings of the 2019 IEEE 39th International Conference on Distributed Computing Systems. Piscataway: IEEE, 2019: 514-524.
[9] MILLER A, RIM K, CHOPRA P, et al. Cooperative perception and localization for cooperative driving[C]//Proceedings of the 2020 IEEE International Conference on Robotics and Automation. Piscataway: IEEE, 2020: 1256-1262.
[10] WANG T H, MANIVASAGAM S, LIANG M, et al. V2VNet: vehicle-to-vehicle communication for joint perception and prediction[C]//Proceedings of the 16th European Conference on Computer Vision. New York: ACM, 2020: 605-621.
[11] XU R S, XIANG H, XIA X, et al. OPV2V: an open benchmark dataset and fusion pipeline for perception with vehicle-to-vehicle communication[C]//Proceedings of the 2022 International Conference on Robotics and Automation. Piscataway: IEEE, 2022: 2583-2589.
[12] HU Y, FANG S H, LEI Z X, et al. Where2comm: communication-efficient collaborative perception via spatial confidence maps[C]//Advances in Neural Information Processing Systems, 2022, 35: 4874-4886.
[13] LANG A H, VORA S, CAESAR H, et al. PointPillars: fast encoders for object detection from point clouds[C]//Proceedings of the 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2019: 12697-12705.
[14] YAN Y, MAO Y X, LI B. SECOND: sparsely embedded convolutional detection[J]. Sensors, 2018, 18(10): 3337.
[15] YANG B, LUO W J, URTASUN R. PIXOR: real-time 3D object detection from point clouds[C]//Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2018: 7652-7660.
[16] ZHOU Y, TUZEL O. VoxelNet: end-to-end learning for point cloud based 3D object detection[C]//Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2018: 4490-4499.
[17] XU R S, XIANG H, TU Z Z, et al. V2X-ViT: vehicle-to-everything cooperative perception withVision transformer[C]//Proceedings of the European Conference on Computer Vision. Cham: Springer, 2022: 107-124.
[18] XU R S, XIA X, LI J L, et al. V2V4Real: a real-world large-scale dataset for vehicle-to-vehicle cooperative perception[C]//Proceedings of the 2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Piscataway: IEEE, 2023: 13712-13722.
[19] XU R S, GUO Y, HAN X, et al. OpenCDA: an open cooperative driving automation framework integrated with co-simulation[C]//Proceedings of the 2021 IEEE International Intelligent Transportation Systems Conference. Piscataway: IEEE, 2021: 1155-1162.
[20] TSUKADA M, OI T, ITO A, et al. AutoC2X: open-source software to realize V2X cooperative perception among autonomous vehicles[C]//Proceedings of the 2020 IEEE 92nd Vehi-
cular Technology Conference. Piscataway: IEEE, 2020: 1-6.