Efficient multi-target tracking via discovering dense subgraphs

Behzad BOZORGTABAR, Roland GOECKE

    Research output: Contribution to journalArticle

    3 Citations (Scopus)
    4 Downloads (Pure)

    Abstract

    In this paper, we cast multi-target tracking as a dense subgraph discovering problem on the undirected relation graph of all given target hypotheses. We aim to extract multiple clusters (dense subgraphs), in which each cluster contains a set of hypotheses of one particular target. In the presence of occlusion or similar moving targets or when there is no reliable evidence for the target's presence, each target trajectory is expected to be fragmented into multiple tracklets. The proposed tracking framework can efficiently link such fragmented target trajectories to build a longer trajectory specifying the true states of the target. In particular, a discriminative scheme is devised via learning the targets' appearance models. Moreover, the smoothness characteristic of the target trajectory is utilised by suggesting a smoothness tracklet affinity model to increase the power of the proposed tracker to produce persistent target trajectories revealing different targets' moving paths. The performance of the proposed approach has been extensively evaluated on challenging public datasets and also in the context of team sports (e.g. soccer, AFL), where team players tend to exhibit quick and unpredictable movements. Systematic experimental results conducted on a large set of sequences show that the proposed approach performs better than the state-of-the-art trackers, in particular, when dealing with occlusion and fragmented target trajectory.
    Original languageEnglish
    Pages (from-to)205-216
    Number of pages12
    JournalComputer Vision and Image Understanding
    Volume144
    Issue numberC
    DOIs
    Publication statusPublished - 2016

    Fingerprint

    Target tracking
    Trajectories
    Sports

    Cite this

    @article{dd69a78c97a74a60871839565e81b23e,
    title = "Efficient multi-target tracking via discovering dense subgraphs",
    abstract = "In this paper, we cast multi-target tracking as a dense subgraph discovering problem on the undirected relation graph of all given target hypotheses. We aim to extract multiple clusters (dense subgraphs), in which each cluster contains a set of hypotheses of one particular target. In the presence of occlusion or similar moving targets or when there is no reliable evidence for the target's presence, each target trajectory is expected to be fragmented into multiple tracklets. The proposed tracking framework can efficiently link such fragmented target trajectories to build a longer trajectory specifying the true states of the target. In particular, a discriminative scheme is devised via learning the targets' appearance models. Moreover, the smoothness characteristic of the target trajectory is utilised by suggesting a smoothness tracklet affinity model to increase the power of the proposed tracker to produce persistent target trajectories revealing different targets' moving paths. The performance of the proposed approach has been extensively evaluated on challenging public datasets and also in the context of team sports (e.g. soccer, AFL), where team players tend to exhibit quick and unpredictable movements. Systematic experimental results conducted on a large set of sequences show that the proposed approach performs better than the state-of-the-art trackers, in particular, when dealing with occlusion and fragmented target trajectory.",
    author = "Behzad BOZORGTABAR and Roland GOECKE",
    year = "2016",
    doi = "10.1016/j.cviu.2015.11.013",
    language = "English",
    volume = "144",
    pages = "205--216",
    journal = "Computer Vision and Image Understanding",
    issn = "1077-3142",
    publisher = "Academic Press Inc.",
    number = "C",

    }

    Efficient multi-target tracking via discovering dense subgraphs. / BOZORGTABAR, Behzad; GOECKE, Roland.

    In: Computer Vision and Image Understanding, Vol. 144, No. C, 2016, p. 205-216.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - Efficient multi-target tracking via discovering dense subgraphs

    AU - BOZORGTABAR, Behzad

    AU - GOECKE, Roland

    PY - 2016

    Y1 - 2016

    N2 - In this paper, we cast multi-target tracking as a dense subgraph discovering problem on the undirected relation graph of all given target hypotheses. We aim to extract multiple clusters (dense subgraphs), in which each cluster contains a set of hypotheses of one particular target. In the presence of occlusion or similar moving targets or when there is no reliable evidence for the target's presence, each target trajectory is expected to be fragmented into multiple tracklets. The proposed tracking framework can efficiently link such fragmented target trajectories to build a longer trajectory specifying the true states of the target. In particular, a discriminative scheme is devised via learning the targets' appearance models. Moreover, the smoothness characteristic of the target trajectory is utilised by suggesting a smoothness tracklet affinity model to increase the power of the proposed tracker to produce persistent target trajectories revealing different targets' moving paths. The performance of the proposed approach has been extensively evaluated on challenging public datasets and also in the context of team sports (e.g. soccer, AFL), where team players tend to exhibit quick and unpredictable movements. Systematic experimental results conducted on a large set of sequences show that the proposed approach performs better than the state-of-the-art trackers, in particular, when dealing with occlusion and fragmented target trajectory.

    AB - In this paper, we cast multi-target tracking as a dense subgraph discovering problem on the undirected relation graph of all given target hypotheses. We aim to extract multiple clusters (dense subgraphs), in which each cluster contains a set of hypotheses of one particular target. In the presence of occlusion or similar moving targets or when there is no reliable evidence for the target's presence, each target trajectory is expected to be fragmented into multiple tracklets. The proposed tracking framework can efficiently link such fragmented target trajectories to build a longer trajectory specifying the true states of the target. In particular, a discriminative scheme is devised via learning the targets' appearance models. Moreover, the smoothness characteristic of the target trajectory is utilised by suggesting a smoothness tracklet affinity model to increase the power of the proposed tracker to produce persistent target trajectories revealing different targets' moving paths. The performance of the proposed approach has been extensively evaluated on challenging public datasets and also in the context of team sports (e.g. soccer, AFL), where team players tend to exhibit quick and unpredictable movements. Systematic experimental results conducted on a large set of sequences show that the proposed approach performs better than the state-of-the-art trackers, in particular, when dealing with occlusion and fragmented target trajectory.

    U2 - 10.1016/j.cviu.2015.11.013

    DO - 10.1016/j.cviu.2015.11.013

    M3 - Article

    VL - 144

    SP - 205

    EP - 216

    JO - Computer Vision and Image Understanding

    JF - Computer Vision and Image Understanding

    SN - 1077-3142

    IS - C

    ER -