Opto-Electronic Engineering, Volume. 50, Issue 12, 230218-1(2024)
Improved CSTrack algorithm for multi-class ship multi-object tracking
Fig. 1. Flowchart of the JDE and CSTrack algorithms. (a) JDE; (b) CSTrack
Fig. 2. Network architecture of the CSTrack. (a) Overall framework; (b) CCN and Res_CCN networks; (c) SAAN network; (d) SAM network; (e) CAM network
Fig. 3. Overall framework and feature extraction network architecture of the proposed method
Fig. 4. Network architecture of the improved Res2net
Fig. 5. Network architecture of CA
Fig. 6. Network architecture of decoupled head
Fig. 7. Flowchart of matching cascade
Fig. 8. Comparison of visualization results between our method and baseline on SMD validation set. (a) FN and FP; (b) ID switch and FN
Fig. 9. Comparison of visualization results between our method and baseline on MOT validation set. (a) FP and FN; (b) ID switch and special FP
|
|
|
|
|
Get Citation
Copy Citation Text
Zhian Yuan, Yu Gu, Gan Ma. Improved CSTrack algorithm for multi-class ship multi-object tracking[J]. Opto-Electronic Engineering, 2024, 50(12): 230218-1
Category: Research Articles
Received: Sep. 1, 2023
Accepted: Dec. 5, 2023
Published Online: Mar. 26, 2024
The Author Email: Gu Yu (谷雨)