• Acta Optica Sinica
  • Vol. 39, Issue 4, 0415003 (2019)
Bin Lin1、2 and Ying Li1、*
Author Affiliations
  • 1 Shaanxi Provincial Key Laboratory of Speech and Image Information Processing, School of Computer Science, Northwestern Polytechnical University, Xi'an, Shaanxi 710129, China
  • 2 School of Science, Guilin University of Technology, Guilin, Guangxi 541004, China
  • show less
    DOI: 10.3788/AOS201939.0415003 Cite this Article Set citation alerts
    Bin Lin, Ying Li. High-Speed Correlation Filter Tracking Algorithm Based on High-Confidence Updating Strategy[J]. Acta Optica Sinica, 2019, 39(4): 0415003 Copy Citation Text show less
    Tracking results and response maps corresponding to proposed algorithm on Jogging-2 sequence. (a) Tracking results; (b) response maps
    Fig. 1. Tracking results and response maps corresponding to proposed algorithm on Jogging-2 sequence. (a) Tracking results; (b) response maps
    Specific embodiment of high-confidence updating strategy in tracking process on Jogging-2 sequence
    Fig. 2. Specific embodiment of high-confidence updating strategy in tracking process on Jogging-2 sequence
    Flowchart of proposed algorithm
    Fig. 3. Flowchart of proposed algorithm
    Tracking results of proposed algorithm with different parameter settings. (a) Precision plot obtained at OPE mode; (b) success plot obtained at OPE mode
    Fig. 4. Tracking results of proposed algorithm with different parameter settings. (a) Precision plot obtained at OPE mode; (b) success plot obtained at OPE mode
    Tracking results of different algorithms on 100 video sequences. (a) Precision plot obtained at OPE mode; (b) success plot obtained at OPE mode; (c) precision plot obtained at SRE mode; (d) success plot obtained at SRE mode; (e) precision plot obtained at TRE mode; (f) success plot obtained at TRE mode
    Fig. 5. Tracking results of different algorithms on 100 video sequences. (a) Precision plot obtained at OPE mode; (b) success plot obtained at OPE mode; (c) precision plot obtained at SRE mode; (d) success plot obtained at SRE mode; (e) precision plot obtained at TRE mode; (f) success plot obtained at TRE mode
    Partial tracking results on eight video sequences. (a) Tiger1; (b) DragonBaby; (c) Bird2; (d) Board; (e) Panda; (f) Jogging-1; (g) Girl2; (h) Human6
    Fig. 6. Partial tracking results on eight video sequences. (a) Tiger1; (b) DragonBaby; (c) Bird2; (d) Board; (e) Panda; (f) Jogging-1; (g) Girl2; (h) Human6
    TrackerIVSVOCCDEFMBFMIPROPROVBCLR
    Proposed66.7064.7364.0760.9966.7663.7563.7863.7557.2765.4964.19
    fDSST68.3962.8258.4656.7264.8264.2367.2461.6053.2971.1359.28
    DSST68.0161.7256.8953.2056.8555.0464.4561.1546.2964.5456.62
    SWCF67.0261.3558.3753.7155.8252.1263.1459.8845.2163.0853.89
    CN54.2851.0751.4450.1645.7046.3060.3857.1442.8457.0647.01
    CFLB37.0544.1641.0239.5639.9240.0945.2941.7733.7438.4455.62
    CSK47.2944.9242.0142.5436.5238.9948.9947.1327.6652.7241.06
    KCF64.2058.5658.1356.8456.4057.5263.2461.7447.9164.5851.14
    Table 1. Precision plot values correspond to different algorithms which are used to test sets of video sequences with different attributes at OPE mode%
    TrackerIVSVOCCDEFMBFMIPROPROVBCLR
    Proposed56.5852.4853.5650.5458.6754.4452.2051.7949.8455.0949.82
    fDSST56.7851.0848.3646.7756.3055.4955.0050.1745.7658.5844.61
    DSST56.1148.5946.1043.4349.2047.1051.0048.2838.4852.4038.94
    SWCF55.2948.0846.7043.4348.3544.8349.9847.1837.8851.0136.82
    CN41.5535.9439.6339.6137.8537.7645.4942.1035.0843.8829.45
    CFLB29.6332.8431.3031.4934.6934.2835.2031.6627.5831.9635.64
    CSK36.8532.3933.1333.7031.3932.6338.0535.3924.9641.0026.33
    KCF47.9239.8644.3043.6245.5644.8447.2245.1239.3349.7730.69
    Table 2. Success plot values correspond to different algorithms which are used to test sets of video sequences with different attributes at OPE mode%
    TrackerScore on precision plot /%Score on success plot /%Average speed /(frame·s-1)
    Proposed (NDR & NHU)64.0152.1637.6
    Proposed (NHU)66.2254.4088.1
    Proposed68.6156.81122.3
    Table 3. Performance evaluation of proposed algorithm in different stages at OPE mode
    TrackerProposedfDSSTDSSTSWCFCNCFLBCSKKCF
    Average speed /(frame·s-1)122.3102.242.622.0239.3190.1523.1291.1
    Table 4. Average tracking speed of different algorithms
    SequenceAttributesFrameObject size /(pixel)
    Tiger1IV, OCC, DEF, MB, FM, IPR, OPR35484×67
    DragonBoySV, OCC, MB, FM, IPR, OPR, OV11365×56
    Bird2OCC, DEF, FM, IPR, OPR9973×69
    BoardSV, MB, FM, OPR, OV, BC698173×198
    PandaSV, OCC, DEF, IPR, OPR, OV, LR100023×28
    Jogging-1OCC, DEF, OPR307101×25
    Girl2SV, OCC, DEF, MB, OPR1500171×44
    Human6SV, OCC, DEF, FM, OPR, OV79255×18
    Table 5. Attributes and relevant information of eight video sequences
    Bin Lin, Ying Li. High-Speed Correlation Filter Tracking Algorithm Based on High-Confidence Updating Strategy[J]. Acta Optica Sinica, 2019, 39(4): 0415003
    Download Citation