Laser & Optoelectronics Progress, Volume. 58, Issue 6, 615003(2021)
A Real-Time Target-Tracking Algorithm Based on Improved SiamFC
Fig. 1. Structure of SiamFC
Fig. 2. Structures of depthwise separable convolution and mixed depthwise convolution
Fig. 3. Network structure
Fig. 4. Success plots of OPE, TRE, SRE on six tracking algorithms. (a) OPE; (b) TRE; (c) SRE
Fig. 5. Precision plots of OPE, TRE, SRE on six tracking algorithms. (a) OPE; (b) TRE; (c) SRE
Fig. 6. Tracking results of our algorithm on OTB2015. (a) David; (b) Bolt; (c) Gym; (d) Car
Fig. 7. Tracking results of our algorithm compared with the other five algorithms in the fast moving video sequences. (a) Video 1; (b) video 2; (c) video 3
Fig. 8. Performance polylines of SiamFC and our algorithm on the ILSVRC2015-train test set. (a) Overlap; (b) accuracy; (c) speed
|
|
|
Get Citation
Copy Citation Text
Zhang Hongying, He Pengyi, Wang Huisan. A Real-Time Target-Tracking Algorithm Based on Improved SiamFC[J]. Laser & Optoelectronics Progress, 2021, 58(6): 615003
Category: Machine Vision
Received: Aug. 11, 2020
Accepted: --
Published Online: Mar. 11, 2021
The Author Email: Hongying Zhang (carole_zhang0716@163.com)