Detail feature pyramid network for small object detection in aerial images

Yiming Xu; Hongbing Ji; Yongquan Zhang; Wenbo Zhang; Peng Wang; Zhaodong Chen

doi:10.1117/12.3025906

8 April 2024 Detail feature pyramid network for small object detection in aerial images

Yiming Xu, Hongbing Ji, Yongquan Zhang, Wenbo Zhang, Peng Wang, Zhaodong Chen

Proceedings Volume 13090, International Conference on Computer Application and Information Security (ICCAIS 2023); 130903I (2024) https://doi.org/10.1117/12.3025906
Event: International Conference on Computer Application and Information Security (ICCAIS 2023), 2023, Wuhan, China

Abstract

Recently, driven by advancements in deep learning, there has been substantial progress in object detection technology. Nonetheless, detecting objects in aerial images remains challenging due to the abundance of small, hard-to-detect objects. Although the multi-scale feature fusion strategy of the feature pyramid network (FPN) broadens the detection scale range, partially alleviating the challenge of small object detection, the cross-scale feature coupling constrains further performance enhancement. To overcome these constraints, we propose a novel Detailed Feature Pyramid Network (D-FPN), specifically designed to improve the efficacy of detecting small objects in aerial images. D-FPN comprises two essential components: the high-resolution feature expansion module (HRFE) and detail information enhancement module (DIE). HRFE extends high-resolution features beneficial for detecting small objects via an asymptotic fusion method. DIE supplements detailed small object information by merging high-resolution features with the backbone’s low-level features. We evaluate D-FPN using the VisDrone2019-DET dataset and compare it with current state-of-the-art object detection methods. Experimental findings validate the effectiveness and advancements achieved by D-FPN.

(2024) Published by SPIE. Downloading of the abstract is permitted for personal use only.

Citation Download Citation

Yiming Xu, Hongbing Ji, Yongquan Zhang, Wenbo Zhang, Peng Wang, and Zhaodong Chen "Detail feature pyramid network for small object detection in aerial images", Proc. SPIE 13090, International Conference on Computer Application and Information Security (ICCAIS 2023), 130903I (8 April 2024); https://doi.org/10.1117/12.3025906

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

Object detection

Education and training

Feature extraction

Feature fusion

Super resolution

Ablation

Image classification

Show All Keywords

Keywords/Phrases

Search In:

Publication Years