A Multi-Object Tracking Approach Combining Contextual Features and Trajectory Prediction

被引:1
|
作者
Zhang, Peng [1 ]
Jing, Qingyang [1 ]
Zhao, Xinlei [2 ]
Dong, Lijia [2 ]
Lei, Weimin [1 ]
Zhang, Wei [1 ]
Lin, Zhaonan [1 ]
机构
[1] Northeastern Univ, Sch Comp Sci & Engn, Shenyang 110169, Peoples R China
[2] Shenyang Er Yi San Elect Technol Co Ltd, Shenyang 110023, Peoples R China
关键词
contextual features; trajectory prediction; trajectory matching; similarity matrix; preprocessing; postprocessing;
D O I
10.3390/electronics12234720
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Aiming to solve the problem of the identity switching of objects with similar appearances in real scenarios, a multi-object tracking approach combining contextual features and trajectory prediction is proposed. This approach integrates the motion and appearance features of objects. The motion features are mainly used for trajectory prediction, and the appearance features are divided into contextual features and individual features, which are mainly used for trajectory matching. In order to accurately distinguish the identities of objects with similar appearances, a context graph is constructed by taking the specified object as the master node and its neighboring objects as the branch nodes. A preprocessing module is applied to exclude unnecessary connections in the graph model based on the speed of the historical trajectory of the object, and to distinguish the features of objects with similar appearances. Feature matching is performed using the Hungarian algorithm, based on the similarity matrix obtained from the features. Post-processing is performed for the temporarily unmatched frames to obtain the final object matching results. The experimental results show that the approach proposed in this paper can achieve the highest MOTA.
引用
收藏
页数:13
相关论文
共 50 条
  • [41] Multi-Object Trajectory Prediction Based on Lane Information and Generative Adversarial Network
    Guo, Lie
    Ge, Pingshu
    Shi, Zhenzhou
    [J]. SENSORS, 2024, 24 (04)
  • [42] Heterogeneous Multi-object Trajectory Prediction Method Based on Hierarchical Graph Attention
    Hu Q.
    Cai Y.
    Wang H.
    Chen L.
    Dong Z.
    Liu Q.
    [J]. Qiche Gongcheng/Automotive Engineering, 2023, 45 (08): : 1448 - 1456
  • [43] MeMOT: Multi-Object Tracking with Memory
    Cai, Jiarui
    Xu, Mingze
    Li, Wei
    Xiong, Yuanjun
    Xia, Wei
    Tu, Zhuowen
    Soatto, Stefano
    [J]. 2022 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2022, : 8080 - 8090
  • [44] Multi-object tracking for horse racing
    Ng, Wing W. Y.
    Liu, Xuyu
    Yan, Xuli
    Tian, Xing
    Zhong, Cankun
    Kwong, Sam
    [J]. INFORMATION SCIENCES, 2023, 638
  • [45] Relational Prior for Multi-Object Tracking
    Moskalev, Artem
    Sosnovik, Ivan
    Smeulders, Arnold
    [J]. 2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION WORKSHOPS (ICCVW 2021), 2021, : 1081 - 1085
  • [46] Multi-Object Tracking with Distributed Sensing
    Dias, Ricardo
    Lau, Nuno
    Silva, Joao
    Lim, Gi Hyun
    [J]. 2016 IEEE INTERNATIONAL CONFERENCE ON MULTISENSOR FUSION AND INTEGRATION FOR INTELLIGENT SYSTEMS (MFI), 2016, : 564 - 569
  • [47] Probability hypothesis density approach for multi-camera multi-object tracking
    Pham, Nam Trung
    Huang, Weimin
    Ong, S. H.
    [J]. COMPUTER VISION - ACCV 2007, PT I, PROCEEDINGS, 2007, 4843 : 875 - +
  • [48] A Robust Framework for Multi-object Tracking
    Jalal, Anand Singh
    Singh, Vrijendra
    [J]. ADVANCES IN COMPUTING AND COMMUNICATIONS, PT 4, 2011, 193 : 329 - 338
  • [49] SiamMOT: Siamese Multi-Object Tracking
    Shuai, Bing
    Berneshawi, Andrew
    Li, Xinyu
    Modolo, Davide
    Tighe, Joseph
    [J]. 2021 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION, CVPR 2021, 2021, : 12367 - 12377
  • [50] HumanTop: a multi-object tracking tabletop
    Soto Candela, Emilio
    Ortega Perez, Mario
    Marin Romero, Clemente
    Perez Lopez, David C.
    Salvador Herranz, Gustavo
    Contero, Manuel
    Alcaniz Raya, Mariano
    [J]. MULTIMEDIA TOOLS AND APPLICATIONS, 2014, 70 (03) : 1837 - 1868