基于多视角低空遥感图像的船舶目标关联

doi:10.7527/S1000-6893.2026.33060

Abstract

Abstract:

Vessel target association under low-altitude remote-sensing scenarios is a crucial component supporting the development of maritime monitoring and intelligent perception systems. However， most existing approaches directly migrate pedestrian or vehicle re-identification algorithms， which fail to effectively handle the unique challenges of vessel imagery-particularly the large intra-class variations and local information loss caused by the diverse imaging perspectives of UAV-based low-altitude imaging platforms. These issues often lead to outlier samples within the same vessel identity， significantly degrading association accuracy. To overcome these limitations， this paper proposes a Multi-scale Correlation-aware Transformer network （MCFormer） for vessel target association. Unlike conventional methods that learn from isolated features of single images， MCFormer performs explicit global and local correlation modeling across multi-scale image collections， leveraging inter-image complementary information to suppress the effects of intra-identity variance and partial occlusion. Specifically， a Global Correlation Module （GCM） constructs a comprehensive inter-image similarity matrix to achieve explicit global correlation modeling through consistency-based feature aggregation， while a Local Correlation Module （LCM） builds a dynamically updated memory bank to mine and align positive local features， capturing fine-grained contextual correlations. Experiments conducted on four publicly available real-world datasets demonstrate that the proposed method consistently outperforms mainstream method in performance metrics related to target association accuracy， verifying its effectiveness， robustness， and engineering potential.

Key words: low-altitude remote sensing, vessel target association, correlation modeling, feature enhancement, feature fusion

CLC Number:

V279

Yunhe LIU, Zhizhuo JIANG, Yu LIU, Xian SUN, You HE. Vessel target association based on multi-view low-altitude remote sensing images[J]. Acta Aeronautica et Astronautica Sinica, 2026, 47(10): 533060.

Figures/Tables 15

Fig.1

Fig.2

Table 1

Table 2

Table 3

Table 4

Table 5

Fig.3

Table 6

Table 7

Fig.4

Fig.5

Table 8

Table 9

Table 10

References 36

[1]	JIA Z Y， ZHU Y A， WU Q H， et al. Remote ID based UAV collision avoidance optimization for low-altitude airspace safety［J］. Chinese Journal of Aeronautics， 2025： 103841.
[2]	REN K， DING L， WAN M J， et al. Target localization based on cross-view matching between UAV and satellite［J］. Chinese Journal of Aeronautics， 2022， 35（9）： 333-341.
[3]	JIN R J， WANG K， LI Z， et al. SAA-O2DINO： Oriented object detection transformer with improved denoising anchor boxes and shape-adaptive assigner［J］. Chinese Journal of Aeronautics， 2025： 103782.
[4]	WANG Y， LI H G， LI X J， et al. UAV image target localization method based on outlier filter and frame buffer［J］. Chinese Journal of Aeronautics， 2024， 37（7）： 375-390.
[5]	WEN Z D， WU J H， LV Y F， et al. Cross-modality vessel re-identification with deep alignment decomposition network［J］. IEEE Transactions on Multimedia， 2024， 26： 10318-10330.
[6]	ZHANG Q， YAN Y M， GAO L， et al. A third-modality collaborative learning approach for visible-infrared vessel reidentification［J］. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing， 2024， 17： 19035-19047.
[7]	LU Z， SUN L G， LV P， et al. A new large-scale dataset for marine vessel re-identification based on Swin transformer network in ocean surveillance scenario［J］. IET Computer Vision， 2025， 19（1）： e70007.
[8]	熊振宇，崔亚奇，董凯，等. 基于属性引导的多源遥感舰船目标可解释融合关联网络［J］. 航空学报， 2023， 44（22）： 627476.
	XIONG Z Y， CUI Y Q， DONG K， et al. Interpretable fusion association network for multi-source remote sensing ship target based on attribute guidance［J］. Acta Aeronautica et Astronautica Sinica， 2023， 44（22）： 627476 （in Chinese）.
[9]	LU F， LAN X Y， ZHANG L J， et al. CricaVPR： Cross-image correlation-aware representation learning for visual place recognition［C］∥2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition （CVPR）. Piscataway： IEEE Press， 2024： 16772-16782.
[10]	ZHANG B， YUAN J K， LI B P， et al. Learning cross-image object semantic relation in transformer for few-shot fine-grained image classification［C］∥Proceedings of the 30th ACM International Conference on Multimedia. New York： ACM， 2022： 2135-2144.
[11]	HAN X R， CHEN Z Q， WANG R X， et al. Joint label refinement and contrastive learning with hybrid memory for unsupervised marine object re-identification［C］∥Proceedings of the 3rd ACM International Conference on Multimedia in Asia. New York： ACM， 2022.
[12]	WANG H， LI S Y， YANG J， et al. Cross-modal ship re-identification via optical and SAR imagery： A novel dataset and method［DB/OL］. arXiv preprint： 2506.22027， 2025.
[13]	DOU W H， ZHU L M， WANG Y， et al. Research on key technology of ship re-identification based on the USV-UAV collaboration［J］. Drones， 2023， 7（9）： 590.
[14]	SPAGNOLO P， FILIERI F， DISTANTE C， et al. A new annotated dataset for boat detection and re-identification［C］∥2019 16th IEEE International Conference on Advanced Video and Signal Based Surveillance （AVSS）. Piscataway： IEEE Press， 2019： 1-7.
[15]	ZENG G M， WANG R J， YU W N， et al. A transfer learning-based approach to maritime warships re-identification［J］. Engineering Applications of Artificial Intelligence， 2023， 125： 106696.
[16]	ZHANG Q， ZHANG M X， LIU J H， et al. Unsupervised maritime vessel re-identification with multi-level contrastive learning［J］. IEEE Transactions on Intelligent Transportation Systems， 2023， 24（5）： 5406-5418.
[17]	XU C A， GAO L， LIU Y， et al. CMShipReID： A cross-modality ship dataset for the reidentification task［J］. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing， 2025， 18： 10503-10513.
[18]	LI Y P， LIU Y Z， ZHANG H Y， et al. Occlusion-aware transformer with second-order attention for person re-identification［J］. IEEE Transactions on Image Processing， 2024， 33： 3200-3211.
[19]	LI Y P， MIAO D Q， ZHANG H Y， et al. Multi-granularity cross transformer network for person re-identification［J］. Pattern Recognition， 2024， 150： 110362.
[20]	ZHENG X T， HUANG X H， JI C， et al. Multi-modal person re-identification based on transformer relational regularization［J］. Information Fusion， 2024， 103： 102128.
[21]	王潇，刘贞报. 基于多层多向Transformer的红外弱小目标检测［J］. 航空学报， 2024， 45（14）： 629490.
	WANG X， LIU Z B. Infrared small target detection based on multi-layer multi-direction transformer［J］. Acta Aeronautica et Astronautica Sinica， 2024， 45（14）： 629490 （in Chinese）.
[22]	刘芳，卢晨阳，路言，等. 基于自适应模板更新的Transformer无人机目标跟踪算法［J］. 航空学报， 2025， 46（16）： 331687.
	LIU F， LU C Y， LU Y， et al. Adaptive template update-based Transformer algorithm for UAV target tracking［J］. Acta Aeronautica et Astronautica Sinica， 2025， 46（16）： 331687 （in Chinese）.
[23]	HE S T， LUO H， WANG P C， et al. TransReID： Transformer-based object re-identification［C］∥2021 IEEE/CVF International Conference on Computer Vision （ICCV）. Piscataway： IEEE Press， 2022： 14993-15002.
[24]	NI H， LI Y K， GAO L L， et al. Part-aware transformer for generalizable person re-identification［C］∥2023 IEEE/CVF International Conference on Computer Vision （ICCV）. Piscataway： IEEE Press， 2024： 11246-11255.
[25]	WANG H C， SHEN J Y， LIU Y T， et al. NFormer： Robust person re-identification with neighbor transformer［C］∥2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition （CVPR）. Piscataway： IEEE Press， 2022： 7287-7297.
[26]	ZHANG Z， ZHANG H J， LIU S. Person re-identification using heterogeneous local graph attention networks［C］∥2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition （CVPR）. Piscataway： IEEE Press， 2021： 12131-12140.
[27]	ZHU H W， KE W J， LI D， et al. Dual cross-attention learning for fine-grained visual categorization and object re-identification［C］∥2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition （CVPR）. Piscataway： IEEE Press， 2022： 4682-4692.
[28]	DOSOVITSKIY A， BEYER L， KOLESNIKOV A， et al. An image is worth 16x16 words： Transformers for image recognition at scale［DB/OL］. arXiv Preprint： 2010.11929， 2020.
[29]	SUN Y F， ZHENG L， YANG Y， et al. Beyond part models： Person retrieval with refined part pooling （and a strong convolutional baseline）［C］∥Computer Vision – ECCV 2018. Cham： Springer， 2018： 501-518.
[30]	WANG T， LIU H， SONG P H， et al. Pose-guided feature disentangling for occluded person re-identification based on transformer［J］. Proceedings of the AAAI Conference on Artificial Intelligence， 2022， 36（3）： 2540-2549.
[31]	CHEN G Y， GU T P， LU J W， et al. Person re-identification via attention pyramid［J］. IEEE Transactions on Image Processing， 2021， 30： 7663-7676.
[32]	DONG N， YAN S L， TANG H， et al. Multi-view information integration and propagation for occluded person re-identification［J］. Information Fusion， 2024， 104： 102201.
[33]	XU H， HUANG L H， CHEN Y T， et al. Unsupervised military-civilian cross-domain vessel re-identification using improved momentum contrast learning［C］∥2024 International Conference on New Trends in Computational Intelligence （NTCI）. Piscataway： IEEE Press， 2024： 106-110.
[34]	ARAIN T ALI， ZHANG P C， MENG Q， et al. Swin transformer with attention mechanism： A novel framework for person re-identification［J］. Pattern Analysis and Applications， 2025， 28（2）： 95.
[35]	WEN Y D， ZHANG K P， LI Z F， et al. A discriminative feature learning approach for deep face recognition［C］∥Computer Vision-ECCV 2016. Cham： Springer， 2016： 499-515.
[36]	DAI Z Z， WANG G Y， YUAN W H， et al. Cluster contrast for unsupervised person re-identification［C］∥Computer Vision-ACCV 2022. Cham： Springer， 2023： 319-337.

数据集	身份数量	图像数量
VesselReID	1 248	30 587
Warships-ReID	163	4 780
BoatRe-ID	107	5 523
AirRGBShipReID	138	3 209

方法	R1/%	R5/%	R10/%	mAP/%
BNN^［29］	63.8	85.3	90.9	50.7
PAT^［24］	66.2	85.7	90.9	53.9
MCL^［16］	63.9	82.1	88.9	45.3
Trans-Reid^［23］	68.2	86.3	91.1	58.7
PFD-Net^［30］	66.1	84.9	90.2	49.2
AP-Net^［31］	62.6	83.3	89.6	50.1
Tran-Aligned^［15］	64.3	82.8	89.4	51.8
MVIIP^［32］	71.2	87.3	91.7	60.8
IMCL^［33］	64.1	82.6	89.5	52.4
SwinReID^［34］	70.7	86.5	92.8	62.3
SwinTransReID^［7］	71.6	89.2	93.3	59.9
MCFormer	72.8	88.9	93.1	63.4

方法	R1/%	R5/%	R10/%	mAP/%
BNN^［29］	87.5	95.8	95.8	70.3
PAT^［24］	91.5	96.3	96.6	75.2
MCL^［16］	86.1	91.1	91.9	56.9
Trans-Reid^［23］	92.1	97.4	97.4	77.1
PFD-Net^［30］	90.5	96.5	96.5	73.9
AP-Net^［31］	82.9	97.1	97.1	53.2
Tran-Aligned^［15］	88.6	97.1	97.1	62.5
MVIIP^［32］	94.2	97.4	97.4	82.7
IMCL^［33］	88.3	95.4	97.1	66.4
SwinReID^［34］	93.7	97.1	97.4	86.6
SwinTransReID^［7］	95.0	98.1	98.1	84.7
MCFormer	96.1	98.1	98.1	88.4

方法	R1/%	mAP/%
BNN^［29］	96.5	89.6
PAT^［24］	97.9	90.6
MCL^［16］	85.0	63.6
Trans-Reid^［23］	98.2	92.1
PFD-Net^［30］	96.3	90.3
AP-Net^［31］	95.9	89.2
Tran-Aligned^［15］	97.5	91.1
MVIIP^［32］	98.4	92.6
IMCL^［33］	95.2	90.5
SwinReID^［34］	98.5	93.2
SwinTransReID^［7］	98.5	93.1
MCFormer	98.8	93.5

方法	R1/%	mAP/%
BNN^［29］	85.6	64.2
PAT^［24］	91.2	72.7
MCL^［16］	81.8	55.4
Trans-Reid^［23］	90.7	72.1
PFD-Net^［30］	88.2	68.4
AP-Net^［31］	86.2	64.7
Tran-Aligned^［15］	88.4	67.9
MVIIP^［32］	91.8	72.7
IMCL^［33］	87.3	66.5
SwinReID^［34］	91.5	72.5
SwinTransReID^［7］	91.8	72.5
MCFormer	92.8	73.2

Vessel target association based on multi-view low-altitude remote sensing images

RichHTML

PDF (PC)

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 15

References 36

Related Articles 15

Recommended Articles

Metrics

Comments

类型	方法	R1/%	mAP/%	Lat/ms	FPS/张
CNN Based	BNN	63.8	50.7	4.9	203.7
	Tran-Aligned	64.3	51.8	6.8	147.1
	MVIIP	71.2	60.8	5.7	175.2
Tansf-ormer Based	Trans-Reid	68.2	58.7	13.6	73.6
	PAT	66.2	53.9	17.6	56.8
	SwinReID	70.7	62.3	16.6	60.2
	SwinTransReID	71.6	59.9	16.3	61.4
本文方法	MCFormer	72.8	63.4	19.2	52.1
	MCFormer*	71.5	62.5	15.3	65.3
	MCFormer+	70.5	60.9	5.2	190.8

模型配置	VesselReID		Warships-ReID
模型配置	R1/%	mAP/%	R1/%	mAP/%
Baseline	66.7	54.5	91.7	77.2
Baseline+GCM	70.2	62.2	95.8	87.3
Baseline+LCM	70.3	57.0	92.9	78.2
完整MCFormer	72.8	63.4	96.1	88.4

损失函数	VesselReID		Warships-ReID
损失函数	R1/%	mAP/%	R1/%	mAP/%
Baseline	66.7	54.5	91.7	77.2
+标准中心损失	68.2	55.8	92.1	77.5
+标准聚类损失	64.5	52.7	89.9	76.1
+局部聚类损失	70.3	57.0	92.9	78.2

局部划分数	VesselReID		Warships-ReID
局部划分数	R1/%	mAP/%	R1/%	mAP/%
2	70.9	61.6	94.8	87.1
3	72.8	63.4	96.1	88.4
4	71.6	62.5	95.6	87.8
5	71.6	62.7	95.6	87.9

融合策略	VesselReID		Warships-ReID
融合策略	R1/%	mAP/%	R1/%	mAP/%
直接相加	69.2	59.5	93.3	84.5
拼接后线性投影	70.1	61.0	95.8	86.2
多尺度通道注意力	72.8	63.4	96.1	88.4

[1]	Zijian XIANG, Zhenyu MA, Xixiang YANG. Inversion of structural performance parameters of composite materials based on deep learning [J]. Acta Aeronautica et Astronautica Sinica, 2025, 46(24): 231877-231877.
[2]	Baoquan SHI, Jiaming ZHOU, Wendong ZHANG, Xianmin CHEN, Qian HE. A network for fatigue crack segmentation and quantification using frequency-domain enhancement and linear attention mechanism [J]. Acta Aeronautica et Astronautica Sinica, 2025, 46(21): 532356-532356.
[3]	Bin SUN, Hang YOU, Wenbo LI, Xiangrui LIU, Jiayi MA. Dual-band payload image fusion and its applications in low-altitude remote sensing [J]. Acta Aeronautica et Astronautica Sinica, 2025, 46(11): 531343-531343.
[4]	Yanfang LIU, Jiayu SHE, Qiufan YUAN, Rui ZHOU, Naiming QI. Real⁃time small target detection networks for UAV remote sensing [J]. Acta Aeronautica et Astronautica Sinica, 2024, 45(14): 630119-630119.
[5]	Aoze YU, Weiwei WEI, Ping WANG, Jinqiang ZHANG, Wenxiong KE. Small target detection algorithm for UAV based on patch⁃wise co⁃attention [J]. Acta Aeronautica et Astronautica Sinica, 2024, 45(14): 629148-629148.
[6]	Zhiqiang FENG, Zhijun XIE, Zhengwei BAO, Kewei CHEN. Real⁃time dense small object detection algorithm for UAV based on improved YOLOv5 [J]. ACTA AERONAUTICAET ASTRONAUTICA SINICA, 2023, 44(7): 327106-327106.
[7]	Guotao MAO, Tianmin DENG, Nanjing YU. Object detection in UAV images based on multi-scale split attention [J]. ACTA AERONAUTICAET ASTRONAUTICA SINICA, 2023, 44(5): 326738-326738.
[8]	Yi ZHANG, Yan ZHANG, Yu ZHANG, Yong ZHANG, Di LIU. Infrared aircraft target detection method based on multi-level feature enhancement fusion [J]. Acta Aeronautica et Astronautica Sinica, 2023, 44(22): 629220-629220.
[9]	Xiaohang LI, Jianjiang ZHOU. Multi⁃scale modality fusion network based on adaptive memory length [J]. Acta Aeronautica et Astronautica Sinica, 2023, 44(22): 628977-628977.
[10]	Yuanliang XUE, Guodong JIN, Lining TAN, Jiankun XU. Adaptive UAV target tracking algorithm based on multi-scale fusion [J]. ACTA AERONAUTICAET ASTRONAUTICA SINICA, 2023, 44(1): 326107-326107.
[11]	LIU Fang, SUN Yanan. UAV target tracking algorithm based on adaptive fusion network [J]. ACTA AERONAUTICAET ASTRONAUTICA SINICA, 2022, 43(7): 325522-325522.
[12]	LIU Fang, HAN Xiao. Adaptive aerial object detection based on multi-scale deep learning [J]. ACTA AERONAUTICAET ASTRONAUTICA SINICA, 2022, 43(5): 325270-325270.
[13]	LIU Chuankai, LI Dongsheng, XIE Jianfeng, LEI Junxiong, YUAN Chunqiang, HE Ximing. Multi-feature fusion based vision locating method for lunar surface sampling teleoperation [J]. ACTA AERONAUTICAET ASTRONAUTICA SINICA, 2022, 43(12): 326296-326296.
[14]	SHI Binbin, CHEN Zhehan. Defect detection method of powder bed based on image feature fusion [J]. ACTA AERONAUTICAET ASTRONAUTICA SINICA, 2021, 42(10): 524430-524430.
[15]	MA Xu, CHENG Yongmei, HAO Shuai, CHEN Kezhe, WANG Tao. Dense point feature generation algorithm based on monocular sequence images for depth measurement of unknown zone [J]. ACTA AERONAUTICAET ASTRONAUTICA SINICA, 2015, 36(2): 596-604.