• Acta Optica Sinica
  • Vol. 38, Issue 5, 0515001 (2018)
Cong Li, Cunyue Lu*, Xun Zhao, Baomin Zhang, and Hongyu Wang
Author Affiliations
  • School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China
  • show less
    DOI: 10.3788/AOS201838.0515001 Cite this Article Set citation alerts
    Cong Li, Cunyue Lu, Xun Zhao, Baomin Zhang, Hongyu Wang. Scale Adaptive Correlation Filtering Tracing Algorithm Based on Feature Fusion[J]. Acta Optica Sinica, 2018, 38(5): 0515001 Copy Citation Text show less
    Tracking process diagram of frame t
    Fig. 1. Tracking process diagram of frame t
    Scale change of object. (a) Carscale; (b) David
    Fig. 2. Scale change of object. (a) Carscale; (b) David
    Multi-scale feature pyramid. (a) Object image; (b) multi-scale image
    Fig. 3. Multi-scale feature pyramid. (a) Object image; (b) multi-scale image
    (a) DP plot; (b) success rate plot
    Fig. 4. (a) DP plot; (b) success rate plot
    Tracking results of different approaches. (a) Basketball; (b) carscale; (c) David; (d) girl; (e) shaking; (f) singer2
    Fig. 5. Tracking results of different approaches. (a) Basketball; (b) carscale; (c) David; (d) girl; (e) shaking; (f) singer2
    DP of algorithms. (a) Basketball; (b) carscale; (c) David; (d) girl; (e) shaking; (f) singer2
    Fig. 6. DP of algorithms. (a) Basketball; (b) carscale; (c) David; (d) girl; (e) shaking; (f) singer2
    Tracking drift phenomenon. (a) Lemming; (b) MotorRolling
    Fig. 7. Tracking drift phenomenon. (a) Lemming; (b) MotorRolling
    Evaluation indexLabRGBYCbCrHSVrgKCF
    Average CLE /pixel19.439.346.324.638.542.7
    Average DP /%81.273.172.676.474.671.8
    Average OP /%74.368.365.671.364.560.5
    Table 1. Comparison of performance of proposed algorithm and KCF algorithm under different color space features
    EvaluationProposedSAMFRPTTGPRDSSTKCF
    Average CLE /pixel19.434.448.556.350.342.7
    Average DP /%81.276.875.170.269.871.8
    Average OP /%74.370.464.663.461.860.8
    Average speed /(frame·s-1)76.018.75.30.741.7258.0
    Table 2. Comparison of performance of proposed algorithm and other algorithms
    SequenceProposedSAMFKCF
    Basketball1.000.970.92
    Carscale0.980.850.80
    David0.930.850.84
    Girl0.950.940.86
    Shaking0.940.910.03
    Singer21.001.000.94
    Table 3. DP of different algorithms
    Cong Li, Cunyue Lu, Xun Zhao, Baomin Zhang, Hongyu Wang. Scale Adaptive Correlation Filtering Tracing Algorithm Based on Feature Fusion[J]. Acta Optica Sinica, 2018, 38(5): 0515001
    Download Citation