• Laser & Optoelectronics Progress
  • Vol. 56, Issue 12, 121501 (2019)
Huan Liu, Chungeng Li*, Jubai An, Guo Wei, and Junli Ren
Author Affiliations
  • Information Science and Technology College, Dalian Maritime University, Dalian, Liaoning 116026, China
  • show less
    DOI: 10.3788/LOP56.121501 Cite this Article Set citation alerts
    Huan Liu, Chungeng Li, Jubai An, Guo Wei, Junli Ren. Multiple Object Tracking Based on Kernelized Correlation Filter[J]. Laser & Optoelectronics Progress, 2019, 56(12): 121501 Copy Citation Text show less
    IOU presentation of ADL-Rundle-6 sequence in 2DMOT2015 dataset
    Fig. 1. IOU presentation of ADL-Rundle-6 sequence in 2DMOT2015 dataset
    Object similarity judgment. (a) 4th frame; (b) 21st frame; (c) template-to-template change
    Fig. 2. Object similarity judgment. (a) 4th frame; (b) 21st frame; (c) template-to-template change
    Status and path of target flow
    Fig. 3. Status and path of target flow
    KCF_IOU tracking algorithm
    Fig. 4. KCF_IOU tracking algorithm
    Effects of different parameters. (a) MOTA; (b) MOTP; (c) FAF; (d) MT; (e) ML; (f) FP; (g) FN; (h) IDS; (i) FM
    Fig. 5. Effects of different parameters. (a) MOTA; (b) MOTP; (c) FAF; (d) MT; (e) ML; (f) FP; (g) FN; (h) IDS; (i) FM
    Comparison among IOU, SORT, and proposed algorithm
    Fig. 6. Comparison among IOU, SORT, and proposed algorithm
    ParameterKlostKsameKscore
    Range2,4,6,810,20,30,400.6,0.7,0.8,0.9
    Table 1. Selectable parameter range of algorithm
    TrackerMOTAMOTPFAFMT /%ML /%FPFNIDSFMHz
    DP-NMS14.570.82.36.040.8131713481445373090444.8
    LDCT4.771.72.411.432.5140663215612348291820.7
    TC_ODAL15.170.52.23.255.9129703853863717161.7
    CEM19.370.72.58.546.5141803459181310231.1
    Proposed29.271.10.910.034.55369365901513285930.0
    Table 2. Performance of proposed approach on 2DMOT2015 sequences
    TrackerMOTAMOTPFAFMT /%ML /%FPFNIDSFMHz
    DP-NMS26.276.30.64.167.536891305573656385.9
    CEM33.275.81.27.854.468371143226427310.3
    EAGS1647.475.91.417.342.7836986931575913197.3
    SMOT29.775.22.95.347.717426107552310844830.2
    MAP-RF64.879.7-37.822.3778057081645-30.0
    Proposed55.876.51.222.833.37046716011856311415.0
    Table 3. Performance of proposed approach on MOT16 sequences
    Huan Liu, Chungeng Li, Jubai An, Guo Wei, Junli Ren. Multiple Object Tracking Based on Kernelized Correlation Filter[J]. Laser & Optoelectronics Progress, 2019, 56(12): 121501
    Download Citation