The number of various types in the DOTA.

<div><p>Ship object detection and fine-grained recognition of remote sensing images are hot topics in remote sensing image processing, with applications in fishing vessel operation command, merchant ship navigation route planning, and other fields. In order to improve the detection accur...

Full description

Saved in:
Bibliographic Details
Main Author: Xuhui Liu (546025) (author)
Other Authors: Chi Feng (7863566) (author), Shuran Zi (22103924) (author), Zhengkun Qin (16850100) (author), Qinghe Guan (22103927) (author)
Published: 2025
Subjects:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1852017402493009920
author Xuhui Liu (546025)
author2 Chi Feng (7863566)
Shuran Zi (22103924)
Zhengkun Qin (16850100)
Qinghe Guan (22103927)
author2_role author
author
author
author
author_facet Xuhui Liu (546025)
Chi Feng (7863566)
Shuran Zi (22103924)
Zhengkun Qin (16850100)
Qinghe Guan (22103927)
author_role author
dc.creator.none.fl_str_mv Xuhui Liu (546025)
Chi Feng (7863566)
Shuran Zi (22103924)
Zhengkun Qin (16850100)
Qinghe Guan (22103927)
dc.date.none.fl_str_mv 2025-08-21T17:35:23Z
dc.identifier.none.fl_str_mv 10.1371/journal.pone.0330485.g008
dc.relation.none.fl_str_mv https://figshare.com/articles/figure/The_number_of_various_types_in_the_DOTA_/29961404
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Neuroscience
Biotechnology
Space Science
Environmental Sciences not elsewhere classified
Biological Sciences not elsewhere classified
Mathematical Sciences not elsewhere classified
Information Systems not elsewhere classified
xlink "> 0
remote sensing images
experimental results show
09 %, respectively
0 visible light
5 </ sub
ship object perception
fuse different scales
spatial position information
div >< p
ship object detection
ship object fine
ship ’
ship objects
different types
different categories
unique features
sub xmlns
sfrm ),
selective memory
regression loss
paper proposes
paper designs
hot topics
grained recognition
grained features
focal loss
detection accuracy
classification loss
based method
dc.title.none.fl_str_mv The number of various types in the DOTA.
dc.type.none.fl_str_mv Image
Figure
info:eu-repo/semantics/publishedVersion
image
description <div><p>Ship object detection and fine-grained recognition of remote sensing images are hot topics in remote sensing image processing, with applications in fishing vessel operation command, merchant ship navigation route planning, and other fields. In order to improve the detection accuracy for different types of remote sensing ship objects, this paper proposes a ship object perception and feature refinement method based on the improved ReDet, called Mamba-ReDet (M-ReDet). First, this paper designs a ship object fine-grained feature extraction backbone (Mamba-ReResNet, M-ReResNet), which selects and reconstructs the unique features of different types of ship objects through the Mamba’s selective memory to improve the algorithm’s ability to extract fine-grained features. Secondly, the M-ReDet consists of the Ship Object Perception Module (SOPM) and the Ship Feature Refinement Module (SFRM), which can extract the ship’s spatial position information from the feature map, fuse different scales of spatial position information and use this information to refine the fine-grained features to improve the detection accuracy of the algorithm for different categories of ships. Finally, we use the KFIoU and Focal Loss as the regression loss and classification loss of the algorithm to improve the accuracy of the training. The experimental results show that the mAP<sub>0.5</sub> of the M-ReDet algorithm on the FAIR1M(ship) and DOTAv1.0 visible light (RGB) remote sensing image datasets are 43.29% and 82.09%, respectively, which is 2.78% and 3.34% higher than that of the ReDet.</p></div>
eu_rights_str_mv openAccess
id Manara_e22cdfd846db5bdaa802b980d010f28b
identifier_str_mv 10.1371/journal.pone.0330485.g008
network_acronym_str Manara
network_name_str ManaraRepo
oai_identifier_str oai:figshare.com:article/29961404
publishDate 2025
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling The number of various types in the DOTA.Xuhui Liu (546025)Chi Feng (7863566)Shuran Zi (22103924)Zhengkun Qin (16850100)Qinghe Guan (22103927)NeuroscienceBiotechnologySpace ScienceEnvironmental Sciences not elsewhere classifiedBiological Sciences not elsewhere classifiedMathematical Sciences not elsewhere classifiedInformation Systems not elsewhere classifiedxlink "> 0remote sensing imagesexperimental results show09 %, respectively0 visible light5 </ subship object perceptionfuse different scalesspatial position informationdiv >< pship object detectionship object fineship ’ship objectsdifferent typesdifferent categoriesunique featuressub xmlnssfrm ),selective memoryregression losspaper proposespaper designshot topicsgrained recognitiongrained featuresfocal lossdetection accuracyclassification lossbased method<div><p>Ship object detection and fine-grained recognition of remote sensing images are hot topics in remote sensing image processing, with applications in fishing vessel operation command, merchant ship navigation route planning, and other fields. In order to improve the detection accuracy for different types of remote sensing ship objects, this paper proposes a ship object perception and feature refinement method based on the improved ReDet, called Mamba-ReDet (M-ReDet). First, this paper designs a ship object fine-grained feature extraction backbone (Mamba-ReResNet, M-ReResNet), which selects and reconstructs the unique features of different types of ship objects through the Mamba’s selective memory to improve the algorithm’s ability to extract fine-grained features. Secondly, the M-ReDet consists of the Ship Object Perception Module (SOPM) and the Ship Feature Refinement Module (SFRM), which can extract the ship’s spatial position information from the feature map, fuse different scales of spatial position information and use this information to refine the fine-grained features to improve the detection accuracy of the algorithm for different categories of ships. Finally, we use the KFIoU and Focal Loss as the regression loss and classification loss of the algorithm to improve the accuracy of the training. The experimental results show that the mAP<sub>0.5</sub> of the M-ReDet algorithm on the FAIR1M(ship) and DOTAv1.0 visible light (RGB) remote sensing image datasets are 43.29% and 82.09%, respectively, which is 2.78% and 3.34% higher than that of the ReDet.</p></div>2025-08-21T17:35:23ZImageFigureinfo:eu-repo/semantics/publishedVersionimage10.1371/journal.pone.0330485.g008https://figshare.com/articles/figure/The_number_of_various_types_in_the_DOTA_/29961404CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/299614042025-08-21T17:35:23Z
spellingShingle The number of various types in the DOTA.
Xuhui Liu (546025)
Neuroscience
Biotechnology
Space Science
Environmental Sciences not elsewhere classified
Biological Sciences not elsewhere classified
Mathematical Sciences not elsewhere classified
Information Systems not elsewhere classified
xlink "> 0
remote sensing images
experimental results show
09 %, respectively
0 visible light
5 </ sub
ship object perception
fuse different scales
spatial position information
div >< p
ship object detection
ship object fine
ship ’
ship objects
different types
different categories
unique features
sub xmlns
sfrm ),
selective memory
regression loss
paper proposes
paper designs
hot topics
grained recognition
grained features
focal loss
detection accuracy
classification loss
based method
status_str publishedVersion
title The number of various types in the DOTA.
title_full The number of various types in the DOTA.
title_fullStr The number of various types in the DOTA.
title_full_unstemmed The number of various types in the DOTA.
title_short The number of various types in the DOTA.
title_sort The number of various types in the DOTA.
topic Neuroscience
Biotechnology
Space Science
Environmental Sciences not elsewhere classified
Biological Sciences not elsewhere classified
Mathematical Sciences not elsewhere classified
Information Systems not elsewhere classified
xlink "> 0
remote sensing images
experimental results show
09 %, respectively
0 visible light
5 </ sub
ship object perception
fuse different scales
spatial position information
div >< p
ship object detection
ship object fine
ship ’
ship objects
different types
different categories
unique features
sub xmlns
sfrm ),
selective memory
regression loss
paper proposes
paper designs
hot topics
grained recognition
grained features
focal loss
detection accuracy
classification loss
based method