航空学报 > 2024, Vol. 45 Issue (7): 329017-329017   doi: 10.7527/S1000-6893.2023.29017

面向反制无人机集群的多目标连续鲁棒跟踪算法

王传云1, 苏阳2, 王琳霖1, 王田3, 王静静4, 高骞1()   

  1. 1.沈阳航空航天大学 人工智能学院,沈阳  110136
    2.沈阳航空航天大学 计算机学院,沈阳  110136
    3.北京航空航天大学 人工智能研究院,北京  100191
    4.中国电子科技集团公司 电子科学研究院,北京  100041
  • 收稿日期:2023-05-19 修回日期:2023-07-20 接受日期:2023-08-17 出版日期:2024-04-15 发布日期:2023-09-01
  • 通讯作者: 高骞 E-mail:gaoqian@buaa.edu.cn
  • 基金资助:
    国家自然科学基金(61703287);辽宁省教育厅科学研究项目(LJKZ0218);沈阳市中青年科技创新人才项目(RC210401);沈阳航空航天大学引进人才科研启动基金(22YB03)

Multi-object continuous robust tracking algorithm for anti-UAV swarm

Chuanyun WANG1, Yang SU2, Linlin WANG1, Tian WANG3, Jingjing WANG4, Qian GAO1()   

  1. 1.College of Artificial Intelligence,Shenyang Aerospace University,Shenyang  110136,China
    2.College of Computer Science,Shenyang Aerospace University,Shenyang  110136,China
    3.Institute of Artificial Intelligence,Beihang University,Beijing  100191,China
    4.China Academy of Electronics and Information Technology,CTEC,Beijing  100041,China
  • Received:2023-05-19 Revised:2023-07-20 Accepted:2023-08-17 Online:2024-04-15 Published:2023-09-01
  • Contact: Qian GAO E-mail:gaoqian@buaa.edu.cn
  • Supported by:
    National Natural Science Foundation of China(61703287);Scientific Research Project of Liaoning Provincial Department of Education(LJKZ0218);Young and middle-aged Science and Technology Innovation Talents Project of Shenyang(RC210401);Doctoral Scientific Research Foundation of Shenyang Aerospace University(22YB03)

摘要:

无人机(UAV)集群作战正朝着智能化、实战化迅猛发展,将在未来战场上造成巨大威胁,面向反制无人机集群的探测与跟踪研究势在必行。针对在复杂场景及远距离探测条件下无人机集群目标之间相互遮挡、无人机为弱小目标等原因造成的检测精度降低和跟踪精度降低问题,本文提出的无人机集群多目标(UAVS-MOT)连续鲁棒跟踪算法可以有效解决。UAVS-MOT模型基于FairMOT模型的多分支无锚框预测结构,将坐标注意力模块与DLA-34网络相结合,构建了全新的主干特征提取网络以提升特征信息的表达能力。此外,引入全新的ArcFace Loss损失函数进行训练以提高模型的收敛速度,并利用BYTE数据关联方法以降低目标漏检率和提高轨迹的连贯性。实验表明,本文提出的UAVS-MOT多目标跟踪算法在UAVSwarm Dataset上的多目标跟踪准确度(MOTA)和目标识别准确度(IDF1)分别为73.4%与76.1%,相比原有FairMOT算法分别提升5.7%与2.9%,可以解决目标的漏检、误检和跟踪精度低的问题,鲁棒性好。

关键词: 多目标跟踪, 无人机集群, FairMOT, 注意力机制, 反制

Abstract:

Unmanned Aerial Vehicle (UAV) swarm warfare is rapidly developing towards intelligence and practicality, which will pose a huge threat on the future battlefield. Therefore, research on detection and tracking of anti-UAV swarm is imperative. To solve the problems of reduced detection and tracking accuracy caused by mutual occlusion between UAV swarm and weak targets in complex scenes and long-distance detection conditions, this paper proposes a UAVS-Multiple Object Tracking (UAVS-MOT) multi-object continuous robust tracking algorithm. The UAVS-MOT model is based on the multi-branch anchor-free frame prediction structure of the FairMOT model. The coordinate attention module is combined with the DLA-34 network to construct a new backbone feature extraction network, so as to improve the expression ability of feature information. In addition, the new ArcFace Loss function is introduced for training to improve the convergence rate of the model, and the BYTE data association method is used to reduce the target miss rate and improve the track consistency. The experiment shows that the Multiple Object Tracking Accuracy (MOTA) and Identity F1 Score (IDF1) of the proposed algorithm on the UAVSwarm Dataset are 73.4% and 76.1%, respectively, which are 5.7% and 2.9% higher than those of the original FairMOT algorithm. The proposed method can solve the problems of missed and false detection of targets and low tracking accuracy, and has good robustness.

Key words: multiple object tracking, UAV swarm, FairMOT, attention mechanism, anti

中图分类号: