UAV small object tracking based on spatio-temporal continuity and feature consistency

Wenqiang Zhan; Qinjie Liu; Jinyan Ma; Xiang Mao; Xu Shu

doi:10.1117/12.2675344

25 May 2023 UAV small object tracking based on spatio-temporal continuity and feature consistency

Wenqiang Zhan, Qinjie Liu, Jinyan Ma, Xiang Mao, Xu Shu

Proceedings Volume 12636, Third International Conference on Machine Learning and Computer Application (ICMLCA 2022); 1263659 (2023) https://doi.org/10.1117/12.2675344
Event: Third International Conference on Machine Learning and Computer Application (ICMLCA 2022), 2022, Shenyang, China

Abstract

Recently, UAVs have gradually been widely used in both military and civilian fields, and play an increasingly important role in various industries. The study of stable and reliable UAV visual target tracking is of great significance to the application of UAVs. Object tracking still faces many challenges, such as target occlusion, target deformation, and other situations. In order to solve the problem that targets are easily lost due to fast moving speed and occlusion, we propose an independent particle based continuous tracking algorithm for the ground object tracking by UAV. We use an iterative update method to get the optimized position estimation of the target at different moments. For accurate and efficient estimation, the features extracted by convolutional neural networks, as well as target trajectory features, are used in the updating process. Multi-scale features are extracted to ensure that the algorithm remains highly accurate for target scale changes. Experimental results show that our algorithm can effectively detect UAV moving targets and still have a good performance for target occlusion.

Citation Download Citation

Wenqiang Zhan, Qinjie Liu, Jinyan Ma, Xiang Mao, and Xu Shu "UAV small object tracking based on spatio-temporal continuity and feature consistency", Proc. SPIE 12636, Third International Conference on Machine Learning and Computer Application (ICMLCA 2022), 1263659 (25 May 2023); https://doi.org/10.1117/12.2675344

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