[1]陆军,王旭东,汲广宇,等.基于恒定转弯率和加速度模型的点云多目标跟踪算法[J].智能系统学报,2025,20(6):1328-1338.[doi:10.11992/tis.202503034]
LU Jun,WANG Xudong,JI Guangyu,et al.Point cloud multitarget tracking algorithm based on the constant turn rate and acceleration model[J].CAAI Transactions on Intelligent Systems,2025,20(6):1328-1338.[doi:10.11992/tis.202503034]
点击复制
LU Jun,WANG Xudong,JI Guangyu,et al.Point cloud multitarget tracking algorithm based on the constant turn rate and acceleration model[J].CAAI Transactions on Intelligent Systems,2025,20(6):1328-1338.[doi:10.11992/tis.202503034]
基于恒定转弯率和加速度模型的点云多目标跟踪算法
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
20
期数:
2025年第6期
页码:
1328-1338
栏目:
学术论文—机器学习
出版日期:
2025-11-05
- Title:
- Point cloud multitarget tracking algorithm based on the constant turn rate and acceleration model
- Keywords:
- point cloud; multitarget tracking; deep learning; autonomous driving; confidence-based two-stage matching; nonlinear dynamics; data association; Hungarian algorithm
- 分类号:
- TP391
- 摘要:
- 针对简单运动模型在复杂驾驶环境多目标跟踪表现不佳的问题，提出了一种基于恒定转弯率和加速度(constant turn rate and acceleration，CTRA)模型的点云多目标跟踪方法。通过采用包含角速度信息的运动模型来描述目标的运动轨迹，可提高在目标转弯时的跟踪精度。同时，利用检测算法提供的速度信息，在轨迹更新时对物体速度进行校正，以改善在目标速度突变时的跟踪效果。此外，采用基于置信度的两阶段匹配策略，以降低低置信度检测框对跟踪结果的影响。在nuScenes验证集上对所提出的三维目标检测与跟踪算法进行了性能评估，并通过消融实验验证了算法中各模块的有效性。实验结果表明，基于CTRA模型的点云多目标跟踪算法在跟踪精度上优于基于简单模型的算法，在目标转弯和速度突变场景下的跟踪效果显著提升，且跟踪过程中身份切换次数大幅减少。
- Abstract:
- To address the suboptimal performance of simple motion models in the context of multitarget tracking within intricate driving environments, this paper proposes a point cloud-based multitarget tracking method that utilizes the constant turn rate and acceleration (CTRA) model. By incorporating angular velocity into the motion model to describe object trajectories, the method improves tracking accuracy during turning maneuvers. Furthermore, the velocity information provided by the detection algorithm is leveraged to correct object speeds during trajectory updates, enhancing robustness under sudden velocity changes. Moreover, a confidence-based two-stage matching strategy is employed to mitigate the influence of low-confidence detections on tracking results. The proposed 3D object detection and tracking algorithm is evaluated on the nuScenes validation set, and ablation studies are conducted to verify the effectiveness of each module. Experimental results demonstrate that the CTRA-based point cloud multitarget tracking algorithm exhibits superior performance in terms of tracking accuracy, significantly enhances performance in turning and abrupt velocity change scenarios, and substantially reduces identity switches during tracking when compared to simple motion models.
- 参考文献/References:
-
[1] LI Bo. 3D fully convolutional network for vehicle detection in point cloud[C]//2017 IEEE/RSJ International Conference on Intelligent Robots and Systems. Vancouver: IEEE, 2017: 1513-1518.
[2] 郭凡, 卢铉宇, 李嘉怡, 等. 基于深度神经网络的目标跟踪算法综述[J]. 航空兵器, 2024, 31(1): 1-12.
GUO Fan, LU Xuanyu, LI Jiayi, et al. A review of target tracking algorithm based on deep neural network[J]. Aero weaponry, 2024, 31(1): 1-12.
[3] 曾雅俊, 王俊, 魏少明, 等. 分布式多传感器多目标跟踪方法综述[J]. 雷达学报, 2023, 12(1): 197-213.
ZENG Yajun, WANG Jun, WEI Shaoming, et al. Review of the method for distributed multi-sensor multi-target tracking[J]. Journal of radars, 2023, 12(1): 197-213.
[4] SIEBEL N T, MAYBANK S. Fusion of multiple tracking algorithms for robust people tracking[C]//European Conference on Computer Vision. Berlin: Springer, 2002: 373-387.
[5] LI Ying, MA Lingfei, ZHONG Zilong, et al. Deep learning for LiDAR point clouds in autonomous driving: a review[J]. IEEE transactions on neural networks and learning systems, 2021, 32(8): 3412-3432.
[6] 孟琭, 杨旭. 目标跟踪算法综述[J]. 自动化学报, 2019, 45(7): 1244-1260.
MENG Lu, YANG Xu. A survey of object tracking algorithms[J]. Acta automatica sinica, 2019, 45(7): 1244-1260.
[7] 杜伟松. 基于机器视觉的智能汽车环境感知算法研究综述[J]. 汽车工业研究, 2018(4): 26-33.
DU Weisong. Overview of research on environment awareness algorithms of smart cars based on machine vision[J]. Auto industry research, 2018(4): 26-33.
[8] 彭湃, 耿可可, 王子威, 等. 智能汽车环境感知方法综述[J]. 机械工程学报, 2023, 59(20): 281-303.
PENG Pai, GENG Keke, WANG Ziwei, et al. Review on environmental perception methods of autonomous vehicles[J]. Journal of mechanical engineering, 2023, 59(20): 281-303.
[9] DU Kehao, BOBKOV A. An overview of object detection and tracking algorithms[C]//INTELS’22. Moscow: MDPI, 2023: 22.
[10] BEWLEY A, GE Zongyuan, OTT L, et al. Simple online and realtime tracking[C]//2016 IEEE International Conference on Image Processing. Phoenix: IEEE, 2016: 3464-3468.
[11] KALMAN R E. A new approach to linear filtering and prediction problems[J]. Journal of basic engineering, 1960, 82(1): 35-45.
[12] REZATOFIGHI H, TSOI N, GWAK J, et al. Generalized intersection over union: a metric and a loss for bounding box regression[C]//2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach: IEEE, 2019: 658-666.
[13] YU Jiahui, JIANG Yuning, WANG Zhangyang, et al. UnitBox: an advanced object detection network[C]//Proceedings of the 24th ACM International Conference on Multimedia. Amsterdam: ACM, 2016: 516-520.
[14] CHIU H K, LI Jie, AMBRU? R, et al. Probabilistic 3D multi-modal, multi-object tracking for autonomous driving[C]//2021 IEEE International Conference on Robotics and Automation. Xi’an: IEEE, 2021: 14227-14233.
[15] MCLACHLAN G J. Mahalanobis distance[J]. Resonance, 1999, 4(6): 20-26.
[16] PANG Ziqi, LI Zhichao, WANG Naiyan. SimpleTrack: understanding and rethinking 3D multi-object tracking[M]//Computer Vision–ECCV 2022 Workshops. Cham: Springer Nature Switzerland, 2023: 680-696.
[17] WANG Sukai, SUN Yuxiang, LIU Chengju, et al. PointTrackNet: an end-to-end network for 3-D object detection and tracking from point clouds[J]. IEEE robotics and automation letters, 2020, 5(2): 3206-3212.
[18] ZHOU Xingyi, KOLTUN V, KR?HENB?HL P. Tracking objects as points[M]//Computer Vision–ECCV 2020. Cham: Springer International Publishing, 2020: 474-490.
[19] 高文印, 文峰, 单铭琦. 基于改进CenterTrack的多目标跟踪算法[J]. 沈阳理工大学学报, 2023, 42(3): 22-27.
GAO Wenyin, WEN Feng, SHAN Mingqi. Multi-target tracking algorithm based on improved CenterTrack[J]. Journal of Shenyang Ligong University, 2023, 42(3): 22-27.
[20] LUO Wenhan, XING Junliang, MILAN A, et al. Multiple object tracking: a literature review[J]. Artificial intelligence, 2021, 293: 103448.
[21] 尹宏鹏, 陈波, 柴毅, 等. 基于视觉的目标检测与跟踪综述[J]. 自动化学报, 2016, 42(10): 1466-1489.
YIN Hongpeng, CHEN Bo, CHAI Yi, et al. Vision-based object detection and tracking: a review[J]. Acta automatica sinica, 2016, 42(10): 1466-1489.
[22] NEUBECK A, VAN GOOL L. Efficient non-maximum suppression[C]//18th International Conference on Pattern Recognition. Hong Kong: IEEE, 2006: 850-855.
[23] 姜佰辰, 关键, 周伟, 等. 基于多项式卡尔曼滤波的船舶轨迹预测算法[J]. 信号处理, 2019, 35(5): 741-746.
JIANG Baichen, GUAN Jian, ZHOU Wei, et al. Vessel trajectory prediction algorithm based on polynomial fitting Kalman filtering[J]. Journal of signal processing, 2019, 35(5): 741-746.
[24] KRENER A J. The convergence of the extended Kalman filter[M]. Berlin: Springer Berlin Heidelberg, 2003.
[25] 张爱民. 扩展卡尔曼滤波在目标跟踪中的应用研究[J]. 信息技术, 2013, 37(10): 95-97,102.
ZHANG Aimin. Research on target tracking of extended Kalman filter and its application[J]. Information technology, 2013, 37(10): 95-97,102.
[26] 郑斌琪, 李宝清, 刘华巍, 等. 采用自适应一致性UKF的分布式目标跟踪[J]. 光学 精密工程, 2019, 27(1): 260-270.
ZHENG Binqi, LI Baoqing, LIU Huawei, et al. Distributed target tracking based on adaptive consensus UKF[J]. Optics and precision engineering, 2019, 27(1): 260-270.
[27] 李天成, 谢昱昕, 李固冲, 等. 面向多目标跟踪的数据关联方法研究综述[J]. 雷达科学与技术, 2025, 23(1): 10-31.
LI Tiancheng, XIE Yuxin, LI Guchong, et al. A survey on data association approaches to multi-target tracking[J]. Radar science and technology, 2025, 23(1): 10-31.
[28] CAESAR H, BANKITI V, LANG A H, et al. nuScenes: a multimodal dataset for autonomous driving[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2020: 11621-11631.
[29] KIM A, O?EP A, LEAL-TAIX? L. EagerMOT: 3D multi-object tracking via sensor fusion[C]//2021 IEEE International Conference on Robotics and Automation. Xi’an: IEEE, 2021: 11315-11321.
[30] BENBARKA N, SCHR?DER J, ZELL A. Score refinement for confidence-based 3D multi-object tracking[C]//2021 IEEE/RSJ International Conference on Intelligent Robots and Systems. Prague: IEEE, 2021: 8083-8090.
- 相似文献/References:
-
[1]伍明,李琳琳,魏振华,等.一种未知环境下机器人多目标跟踪算法[J].智能系统学报,2015,10(3):448.[doi:10.3969/j.issn.1673-4785.201405051]
WU Ming,LI Linlin,WEI Zhenhua,et al.A robot multi-object tracking algorithm in unknown environments[J].CAAI Transactions on Intelligent Systems,2015,10():448.[doi:10.3969/j.issn.1673-4785.201405051]
[2]张金艺,梁滨,唐笛恺,等.粗匹配和局部尺度压缩搜索下的快速ICP-SLAM[J].智能系统学报,2017,12(3):413.[doi:10.11992/tis.201605029]
ZHANG Jinyi,LIANG Bin,TANG Dikai,et al.Fast ICP-SLAM with rough alignment and local scale-compressed searching[J].CAAI Transactions on Intelligent Systems,2017,12():413.[doi:10.11992/tis.201605029]
[3]欧伟奇,尹辉,许宏丽,等.一种基于Multi-Egocentric视频运动轨迹重建的多目标跟踪算法[J].智能系统学报,2019,14(2):246.[doi:10.11992/tis.201709003]
OU Weiqi,YIN Hui,XU Hongli,et al.A multi-object tracking algorithm based on trajectory reconstruction on multi-egocentric video[J].CAAI Transactions on Intelligent Systems,2019,14():246.[doi:10.11992/tis.201709003]
[4]鲁斌,孙洋,杨振宇.融合体素图注意力的三维目标检测算法[J].智能系统学报,2024,19(3):598.[doi:10.11992/tis.202209008]
LU Bin,SUN Yang,YANG Zhenyu.3D object detection algorithm with voxel graph attention[J].CAAI Transactions on Intelligent Systems,2024,19():598.[doi:10.11992/tis.202209008]
[5]陆军,鲁林超,翟晓阳,等.面向道路交通场景的高效3D目标检测[J].智能系统学报,2025,20(1):91.[doi:10.11992/tis.202311013]
LU Jun,LU Linchao,ZHAI Xiaoyang,et al.High-efficiency 3D object detection for road traffic scenes[J].CAAI Transactions on Intelligent Systems,2025,20():91.[doi:10.11992/tis.202311013]
备注/Memo
收稿日期:2025-3-24。
基金项目:黑龙江省自然科学基金项目（F201123）.
作者简介:陆军，教授，博士生导师，主要研究方向为计算机视觉、机器感知、 机械臂控制。科技部科技型中小企业创新基金项目评审专家，国家自然科学基金同行评议专家。编写著作5部，发表学术论文80余篇。E-mail： lujun0260@sina.com。;王旭东，硕士研究生，主要研究方向为计算机视觉。E-mail：wangxudongheu@163.com。;汲广宇，博士研究生，主要研究方向为计算机视觉、深度学习以及基于点云的三维目标检测。E-mail：jiguangyu94@163.com。
通讯作者:陆军. E-mail：lujun0260@sina.com
更新日期/Last Update:
1900-01-01