Shape-Dependent Dynamic Label Assignment for Oriented Remote Sensing Object Detection

Oriented remote sensing object detection (ORSOD) has gained increasing significance in both military and civilian applications due to the necessity of accurately identifying objects with varying shapes and orientations in remote sensing data. Traditional ORSOD methods often employ fixed label assign...

Full description

Saved in:
Bibliographic Details
Main Authors: Xue Zhang, Yanxia Wu, Guoyin Zhang, Ye Yuan, Guangliang Cheng, Yulei Wu
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10745646/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850219367685423104
author Xue Zhang
Yanxia Wu
Guoyin Zhang
Ye Yuan
Guangliang Cheng
Yulei Wu
author_facet Xue Zhang
Yanxia Wu
Guoyin Zhang
Ye Yuan
Guangliang Cheng
Yulei Wu
author_sort Xue Zhang
collection DOAJ
description Oriented remote sensing object detection (ORSOD) has gained increasing significance in both military and civilian applications due to the necessity of accurately identifying objects with varying shapes and orientations in remote sensing data. Traditional ORSOD methods often employ fixed label assignment strategies to differentiate between positive and negative samples. However, most of them frequently overlook the impact of object shape on sample quality, leading to an imbalanced distribution of positive samples and exacerbating the inconsistency between classification and regression tasks, thereby limiting detection performance. To address these challenges, we propose a novel shape-dependent assignment (SDA) method that dynamically differentiates positive and negative samples based on object shape. It introduces a new metric for evaluating sample box quality by considering angular differences relative to ground truth (GT) boxes and adjusts the sample scoring threshold according to the aspect ratio of each GT box. In addition, we present a DIoU-adaptive weighting (DAW) module that enhances the interaction between classification and regression tasks by leveraging the distance-IoU metric. This approach not only balances the quantity of samples but also improves their quality, enabling more effective training schemes for samples of varying qualities. We validate our proposed methods through extensive experiments on three challenging ORSOD datasets: DOTA-1.0, HRSC2016, and UCAS-AOD. The results demonstrate that our approach achieves significant improvements, especially for objects with large aspect ratios.
format Article
id doaj-art-7e70fa480de6470da4aa1951af7e21e8
institution OA Journals
issn 1939-1404
2151-1535
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
spelling doaj-art-7e70fa480de6470da4aa1951af7e21e82025-08-20T02:07:24ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352025-01-011813214610.1109/JSTARS.2024.349216410745646Shape-Dependent Dynamic Label Assignment for Oriented Remote Sensing Object DetectionXue Zhang0https://orcid.org/0009-0004-7567-8614Yanxia Wu1https://orcid.org/0000-0001-8384-9234Guoyin Zhang2https://orcid.org/0000-0001-6925-3039Ye Yuan3https://orcid.org/0000-0002-8210-5054Guangliang Cheng4https://orcid.org/0000-0002-1428-8848Yulei Wu5https://orcid.org/0000-0003-0801-8443College of Computer Science and Technology, Harbin Engineering University, Harbin, ChinaCollege of Computer Science and Technology, Harbin Engineering University, Harbin, ChinaCollege of Computer Science and Technology, Harbin Engineering University, Harbin, ChinaCollege of Computer Science and Technology, Harbin Engineering University, Harbin, ChinaDepartment of Computer Science, University of Liverpool, Liverpool, U.K.Faculty of Science and Engineering, University of Bristol, Bristol, U.K.Oriented remote sensing object detection (ORSOD) has gained increasing significance in both military and civilian applications due to the necessity of accurately identifying objects with varying shapes and orientations in remote sensing data. Traditional ORSOD methods often employ fixed label assignment strategies to differentiate between positive and negative samples. However, most of them frequently overlook the impact of object shape on sample quality, leading to an imbalanced distribution of positive samples and exacerbating the inconsistency between classification and regression tasks, thereby limiting detection performance. To address these challenges, we propose a novel shape-dependent assignment (SDA) method that dynamically differentiates positive and negative samples based on object shape. It introduces a new metric for evaluating sample box quality by considering angular differences relative to ground truth (GT) boxes and adjusts the sample scoring threshold according to the aspect ratio of each GT box. In addition, we present a DIoU-adaptive weighting (DAW) module that enhances the interaction between classification and regression tasks by leveraging the distance-IoU metric. This approach not only balances the quantity of samples but also improves their quality, enabling more effective training schemes for samples of varying qualities. We validate our proposed methods through extensive experiments on three challenging ORSOD datasets: DOTA-1.0, HRSC2016, and UCAS-AOD. The results demonstrate that our approach achieves significant improvements, especially for objects with large aspect ratios.https://ieeexplore.ieee.org/document/10745646/High-quality sampleslabel assignmentobjects with large aspect ratiosoriented remote sensing object detection (ORSOD)
spellingShingle Xue Zhang
Yanxia Wu
Guoyin Zhang
Ye Yuan
Guangliang Cheng
Yulei Wu
Shape-Dependent Dynamic Label Assignment for Oriented Remote Sensing Object Detection
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
High-quality samples
label assignment
objects with large aspect ratios
oriented remote sensing object detection (ORSOD)
title Shape-Dependent Dynamic Label Assignment for Oriented Remote Sensing Object Detection
title_full Shape-Dependent Dynamic Label Assignment for Oriented Remote Sensing Object Detection
title_fullStr Shape-Dependent Dynamic Label Assignment for Oriented Remote Sensing Object Detection
title_full_unstemmed Shape-Dependent Dynamic Label Assignment for Oriented Remote Sensing Object Detection
title_short Shape-Dependent Dynamic Label Assignment for Oriented Remote Sensing Object Detection
title_sort shape dependent dynamic label assignment for oriented remote sensing object detection
topic High-quality samples
label assignment
objects with large aspect ratios
oriented remote sensing object detection (ORSOD)
url https://ieeexplore.ieee.org/document/10745646/
work_keys_str_mv AT xuezhang shapedependentdynamiclabelassignmentfororientedremotesensingobjectdetection
AT yanxiawu shapedependentdynamiclabelassignmentfororientedremotesensingobjectdetection
AT guoyinzhang shapedependentdynamiclabelassignmentfororientedremotesensingobjectdetection
AT yeyuan shapedependentdynamiclabelassignmentfororientedremotesensingobjectdetection
AT guangliangcheng shapedependentdynamiclabelassignmentfororientedremotesensingobjectdetection
AT yuleiwu shapedependentdynamiclabelassignmentfororientedremotesensingobjectdetection