Paper
9 October 2023 Deformable convolution-based motion target detection algorithm
Yijun Tang, Hong Fan, Ruyi Sun, Yi Yang, Shuyu Yu
Author Affiliations +
Proceedings Volume 12791, Third International Conference on Advanced Algorithms and Neural Networks (AANN 2023); 1279108 (2023) https://doi.org/10.1117/12.3004658
Event: Third International Conference on Advanced Algorithms and Neural Networks (AANN 2023), 2023, Qingdao, SD, China
Abstract
For motion target detection in dynamic backgrounds, most traditional methods have drawbacks, such as long computation time or high limitation on the background. This paper proposes a deformable convolution-based motion target detection algorithm by replacing part of the C3 module in the YOLOv5 feature extraction layer with deformable convolution (DCNv3) to introduce long distance dependence and adaptive spatial aggregation, and adding an ECA attention mechanism to reduce the effect of background variation. The the accuracy, recall and mAP of the improved YOLOv5s algorithm increases by 1.6%, 3.8% and 2.3% respectively, and is able to identify moving targets in dynamic backgrounds more accurately than the original algorithm.
(2023) Published by SPIE. Downloading of the abstract is permitted for personal use only.
Yijun Tang, Hong Fan, Ruyi Sun, Yi Yang, and Shuyu Yu "Deformable convolution-based motion target detection algorithm", Proc. SPIE 12791, Third International Conference on Advanced Algorithms and Neural Networks (AANN 2023), 1279108 (9 October 2023); https://doi.org/10.1117/12.3004658
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Target detection

Convolution

Deformation

Animal model studies

Animals

Motion detection

Motion models

Back to Top