Multi-target tracking based multi-level matching algorithm

Yang Geng; Changzhen Xiong; Jiawei Chen

doi:10.1117/12.2639856

19 October 2022 Multi-target tracking based multi-level matching algorithm

Yang Geng, Changzhen Xiong, Jiawei Chen

Proceedings Volume 12294, 7th International Symposium on Advances in Electrical, Electronics, and Computer Engineering; 1229443 (2022) https://doi.org/10.1117/12.2639856
Event: 7th International Symposium on Advances in Electrical, Electronics and Computer Engineering (ISAEECE 2022), 2022, Xishuangbanna, China

Abstract

In order to solve the problem that most tracking algorithms' re-identification models learn the features of target detection directly, resulting in low quality of generated re-identification features; and the problem that the number of target label switching increases due to occlusion and loss between multiple targets. In this paper, a multi-target tracking algorithm based on multi-level matching is proposed. The algorithm uses a shared structure, which effectively reduces the amount of redundant computation and can achieve almost real-time tracking. The algorithm also uses lower-order appearance features for re-identification operations based on the property that appearance features are more likely to distinguish different instances of the same class, effectively reducing the number of target switches. Finally, a multi-level matching strategy using cascade matching, IOU matching and re-identification feature matching in the association phase can effectively target targets where occlusion occurs and improve tracking accuracy. The algorithm was validated on the MOT2015 standard dataset, and the experimental results showed that the Multiple Object Tracking Accuracy (MOTA) of the algorithm could reach 66.5%; the number of target label switches (IDs) was reduced to 375. The algorithm in this paper effectively reduces the number of target label switches and improves the tracking accuracy. Compared with the other five groups of advanced tracking algorithms in the experiment, the algorithm in this paper has a large improvement in performance.

Citation Download Citation

Yang Geng, Changzhen Xiong, and Jiawei Chen "Multi-target tracking based multi-level matching algorithm", Proc. SPIE 12294, 7th International Symposium on Advances in Electrical, Electronics, and Computer Engineering, 1229443 (19 October 2022); https://doi.org/10.1117/12.2639856

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available

Members: $17.00

Non-members: $21.00 ADD TO CART

PROCEEDINGS
6 PAGES

DOWNLOAD PAPER SAVE TO MY LIBRARY

GET CITATION

RIGHTS & PERMISSIONS

Get copyright permission Get copyright permission on Copyright Marketplace

KEYWORDS

Detection and tracking algorithms

Target detection

Optical tracking

Filtering (signal processing)

Switching

Visualization

Image processing

Show All Keywords

Keywords/Phrases

Search In:

Publication Years