RSW-YOLO: A Vehicle Detection Model for Urban UAV Remote Sensing Images

Vehicle detection in remote sensing images faces significant challenges due to small object sizes, scale variation, and cluttered backgrounds. To address these issues, we propose RSW-YOLO, an enhanced detection model built upon the YOLOv8n framework, designed to improve feature extraction and robust...

Full description

Saved in:
Bibliographic Details
Main Authors: Hao Wang, Jiapeng Shang, Xinbo Wang, Qingqi Zhang, Xiaoli Wang, Jie Li, Yan Wang
Format: Article
Language:English
Published: MDPI AG 2025-07-01
Series:Sensors
Subjects:
Online Access:https://www.mdpi.com/1424-8220/25/14/4335
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849418018857156608
author Hao Wang
Jiapeng Shang
Xinbo Wang
Qingqi Zhang
Xiaoli Wang
Jie Li
Yan Wang
author_facet Hao Wang
Jiapeng Shang
Xinbo Wang
Qingqi Zhang
Xiaoli Wang
Jie Li
Yan Wang
author_sort Hao Wang
collection DOAJ
description Vehicle detection in remote sensing images faces significant challenges due to small object sizes, scale variation, and cluttered backgrounds. To address these issues, we propose RSW-YOLO, an enhanced detection model built upon the YOLOv8n framework, designed to improve feature extraction and robustness against environmental noise. A Restormer module is incorporated into the backbone to model long-range dependencies via self-attention, enabling better handling of multi-scale features and complex scenes. A dedicated detection head is introduced for small objects, focusing on critical channels while suppressing irrelevant information. Additionally, the original CIoU loss is replaced with WIoU, which dynamically reweights predicted boxes based on their quality, enhancing localization accuracy and stability. Experimental results on the DJCAR dataset show mAP@0.5 and mAP@0.5:0.95 improvements of 5.4% and 6.2%, respectively, and corresponding gains of 4.3% and 2.6% on the VisDrone dataset. These results demonstrate that RSW-YOLO offers a robust and accurate solution for UAV-based vehicle detection, particularly in urban scenes with dense or small targets.
format Article
id doaj-art-d0c56d3a431f4061a7b956c0c01b77fc
institution Kabale University
issn 1424-8220
language English
publishDate 2025-07-01
publisher MDPI AG
record_format Article
series Sensors
spelling doaj-art-d0c56d3a431f4061a7b956c0c01b77fc2025-08-20T03:32:33ZengMDPI AGSensors1424-82202025-07-012514433510.3390/s25144335RSW-YOLO: A Vehicle Detection Model for Urban UAV Remote Sensing ImagesHao Wang0Jiapeng Shang1Xinbo Wang2Qingqi Zhang3Xiaoli Wang4Jie Li5Yan Wang6Electronic Information Engineering College, Changchun University, Changchun 130022, ChinaElectronic Information Engineering College, Changchun University, Changchun 130022, ChinaElectronic Information Engineering College, Changchun University, Changchun 130022, ChinaElectronic Information Engineering College, Changchun University, Changchun 130022, ChinaElectronic Information Engineering College, Changchun University, Changchun 130022, ChinaElectronic Information Engineering College, Changchun University, Changchun 130022, ChinaElectronic Information Engineering College, Changchun University, Changchun 130022, ChinaVehicle detection in remote sensing images faces significant challenges due to small object sizes, scale variation, and cluttered backgrounds. To address these issues, we propose RSW-YOLO, an enhanced detection model built upon the YOLOv8n framework, designed to improve feature extraction and robustness against environmental noise. A Restormer module is incorporated into the backbone to model long-range dependencies via self-attention, enabling better handling of multi-scale features and complex scenes. A dedicated detection head is introduced for small objects, focusing on critical channels while suppressing irrelevant information. Additionally, the original CIoU loss is replaced with WIoU, which dynamically reweights predicted boxes based on their quality, enhancing localization accuracy and stability. Experimental results on the DJCAR dataset show mAP@0.5 and mAP@0.5:0.95 improvements of 5.4% and 6.2%, respectively, and corresponding gains of 4.3% and 2.6% on the VisDrone dataset. These results demonstrate that RSW-YOLO offers a robust and accurate solution for UAV-based vehicle detection, particularly in urban scenes with dense or small targets.https://www.mdpi.com/1424-8220/25/14/4335deep learningremote sensing imagesobject detectionYOLO
spellingShingle Hao Wang
Jiapeng Shang
Xinbo Wang
Qingqi Zhang
Xiaoli Wang
Jie Li
Yan Wang
RSW-YOLO: A Vehicle Detection Model for Urban UAV Remote Sensing Images
Sensors
deep learning
remote sensing images
object detection
YOLO
title RSW-YOLO: A Vehicle Detection Model for Urban UAV Remote Sensing Images
title_full RSW-YOLO: A Vehicle Detection Model for Urban UAV Remote Sensing Images
title_fullStr RSW-YOLO: A Vehicle Detection Model for Urban UAV Remote Sensing Images
title_full_unstemmed RSW-YOLO: A Vehicle Detection Model for Urban UAV Remote Sensing Images
title_short RSW-YOLO: A Vehicle Detection Model for Urban UAV Remote Sensing Images
title_sort rsw yolo a vehicle detection model for urban uav remote sensing images
topic deep learning
remote sensing images
object detection
YOLO
url https://www.mdpi.com/1424-8220/25/14/4335
work_keys_str_mv AT haowang rswyoloavehicledetectionmodelforurbanuavremotesensingimages
AT jiapengshang rswyoloavehicledetectionmodelforurbanuavremotesensingimages
AT xinbowang rswyoloavehicledetectionmodelforurbanuavremotesensingimages
AT qingqizhang rswyoloavehicledetectionmodelforurbanuavremotesensingimages
AT xiaoliwang rswyoloavehicledetectionmodelforurbanuavremotesensingimages
AT jieli rswyoloavehicledetectionmodelforurbanuavremotesensingimages
AT yanwang rswyoloavehicledetectionmodelforurbanuavremotesensingimages