UAV Geolocation based on Spatial Information-Enhanced Siamese Network

DENG Zhen; HUI Bingwei

doi:10.12382/bgxb.2026.0380

您当前的位置：

首页 >

文章列表页 >

UAV Geolocation based on Spatial Information-Enhanced Siamese Network

更新时间：2026-05-19

- UAV Geolocation based on Spatial Information-Enhanced Siamese Network
- Acta Armamentarii (2026)
- 作者机构：
  
  国防科技大学电子科学学院,湖南,长沙,410000
- 作者简介：
- 基金信息：
- DOI：10.12382/bgxb.2026.0380
  CLC： TP391.41
- Received：27 April 2026，
  
  Online First：19 May 2026，
- 稿件说明：
移动端阅览
邓臻，回丙伟. 基于空间信息增强孪生网络的无人机地理定位[J/OL]. 兵工学报, 2026(2026-05-19). https://doi.org/10.12382/bgxb.2026.0380.

DENG Z, HUI B W. Uav geolocation based on spatial information-enhanced siamese network[J/OL]. Acta Armamentarii, 2026(2026-05-19). https://doi.org/10.12382/bgxb.2026.0380. (in Chinese)
邓臻，回丙伟. 基于空间信息增强孪生网络的无人机地理定位[J/OL]. 兵工学报, 2026(2026-05-19). https://doi.org/10.12382/bgxb.2026.0380. DOI：

DENG Z, HUI B W. Uav geolocation based on spatial information-enhanced siamese network[J/OL]. Acta Armamentarii, 2026(2026-05-19). https://doi.org/10.12382/bgxb.2026.0380. (in Chinese) DOI：

摘要

在卫星导航受限的环境中，基于无人机实时图像与卫星参考图像匹配的地理定位性能受限于这两类图像的视角差异与场景复杂性。针对此问题，提出基于空间信息增强孪生网络的无人机地理定位方法：设计多尺度模块来同步捕获全局语义以及局部细节，提高模型的复杂场景适配性；引入坐标注意力机制，设计双分支坐标注意力模块以建模长距离依赖，并聚焦关键特征区域，进而提升模型对显著特征的辨识能力。在GTA-UAV数据集上的实验表明：same-area 场景中，召回率R@1达到86.24%；cross-area场景下，R@1达到58.67%，充分验证了所提方法的有效性。

Abstract

In satellite navigation-constrained environments

the geolocation performance based on matching UAV real-time images with satellite reference images is constrained by the viewing angle differences and scene complexity between these two types of images. To address this issue

this paper proposes a UAV geolocation method based on a spatial information-enhancedsiamesenetwork: on the one hand

a multi-scale module is designed to simultaneously capture global semantic information and local detailed features

improving the model's adaptability to complex scenes; on the other hand

the coordinate attention mechanism is introduced

and a dual-branch coordinate attention module is designed to model long-range dependencies and focus on key feature regions

thereby enhancing the model's ability to identify salient features. Experiments on the GTA-UAV dataset show that in the same-area scenario

the recall rate R@1 reaches 86.24%; in the cross-area scenario

R@1 reaches 58.67%

which fully verifies the effectiveness of the proposed method.

关键词

Keywords

references

廖小罕, 屈文秋, 徐晨晨, 等. 城市空中交通及其新型基础设施低空公共航路研究综述[J]. 航空学报, 2023, 44(24): 6-34.

LIAO X H, QU W Q, XU C C, et al. A review of urban air mobility and its new infrastructure low-altitude public routes[J]. Acta Aeronautica et Astronautica Sinica, 2023, 44(24): 6-34.

HE M, CHEN C, LIU J, et al. AerialVL: A dataset, baseline and algorithm framework for aerial-based visual localization with reference map[J]. IEEE Robotics and Automation Letters, 2024, 9(10): 8210-8217.

XIAO J, ZHANG N, TORTEI D, et al. STHN: Deep homography estimation for UAV thermal geo-localization with satellite imagery[J]. IEEE Robotics and Automation Letters, 2024, 9(10): 1-8.

BANSAL M, SAWHNEY H S, CHENG H, et al. Geo-localization of street views with aerial image databases[C]// Proceedings of the 19th ACM International Conference on Multimedia. Scottsdale, USA: ACM, 2011: 1125-1128.

CASTALDO F, ZAMIR A, ANGST R, et al. Semantic cross-view matching[C]//Proceedings of the 15th IEEE International Conference on Computer Vision Workshops. Santiago, Chile: IEEE, 2015: 9-17.

WORKMAN S, JACOBS N, et al. On the location dependence of convolutional neural network features[C]//Proceedings of the 28th IEEE Conference on Computer Vision and Pattern Recognition Workshops. Boston, USA: IEEE, 2015: 70-78.

WORKMAN S, SOUVENIR R, JACOBS N. Wide-area image geolocalization with aerial reference imagery[C]//Proceedings of the 15th IEEE International Conference on Computer Vision. Santiago, Chile: IEEE, 2015: 3961-3969.

CHOPRA S, HADSELL R, LECUN Y. Learning a similarity metric discriminatively, with application to face verification[C]//Proceedings of the 18th IEEE Computer Society Conference on Computer Vision and Pattern Recognition. San Diego, USA: IEEE, 2005: 539-546.

LIN T Y, CUI Y, BELONGIE S, et al. Learning deep representations for ground-to-aerial geolocalization[C]//Proceedings of the 28th IEEE Conference on Computer Vision and Pattern Recognition. Boston, USA: IEEE, 2015: 5007-5015.

VO N N, HAYS J. Localizing and orienting street views using overhead imagery[C]//Proceedings of the 14th European Conference on Computer Vision. Cham, Switzerland: Springer, 2016: 494-509.

ZHENG Z, WEI Y, YANG Y. University-1652: a multi-view multi-source benchmark for drone-based geo-localization[C]//Proceedings of the 28th ACM international conference on Multimedia. 2020: 1395-1403.

ZHENG Z, WEI Y, YANG Y. University-1652: A multi-view multi-source benchmark for drone-based geo-localization[C]//Proceedings of the 20th ACM International Conference on Multimedia. Seattle, USA: ACM, 2020: 1395-1403.

ZHENG Z, ZHENG L, YANG Y. Unlabeled samples generated by GAN improve the person re-identification baseline in vitro[C]//Proceedings of the 16th IEEE International Conference on Computer Vision. Venice, Italy: IEEE, 2017: 3754-3762.

WANG T, ZHENG Z, SUN Y, et al. Multiple-environment self-adaptive network for aerial-view geo-localization[J]. Pattern Recognition, 2024, 152: 110363.

YANG H, LU X, ZHU Y. Cross-view geo-localization with layer-to-layer transformer[J]. Advances in Neural Information Processing Systems, 2021, 34: 29009-29020.

DAI M, ZHENG E, FENG Z, et al. Vision-based UAV self-positioning in low-altitude urban environments[J]. IEEE Transactions on Image Processing, 2023, 33: 493-508.

ZHU R, YIN L, YANG M, et al. SUES-200: A multi-height multi-scene cross-view image benchmark across drone and satellite[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2023, 33(9): 4825-4839.

DEUSER F, HABEL K, OSWALD N. Sample4geo: Hard negative sampling for cross-view geo-localisation[C]//Proceedings of the 19th IEEE/CVF International Conference on Computer Vision. Paris, France: IEEE, 2023: 16847-16856.

RADFORD A, KIM J W, HALLACY C, et al. Learning transferable visual models from natural language supervision[C]//Proceedings of the 38th International Conference on Machine Learning. [S.l.]: PMLR, 2021: 8748-8763.

XU W, YAO Y, CAO J, et al. UAV-visloc: A large-scale dataset for UAV visual localization[J]. arXiv Preprint arXiv:2405.11936, 2024.

JI Y, HE B, TAN Z, et al. Game4Loc: A UAV geo-localization benchmark from game data[C]//Proceedings of the 39th AAAI Conference on Artificial Intelligence. Philadelphia, USA: AAAI, 2025: 3913-3921.

Views

下载量

CNKI被引量

Alert me when the article has been cited

提交

Tools

Publicity Resources

Guidance-Tracker:An Adaptive UAV Siamese Tracker for Visual Guidance

Research on the Identification of Spherical Explosive Fragmentation Damage Effect Based on Siamese Networks and Regional Attention Mechanisms

Lightweight and Fast Target Tracking Algorithm Based on Ghost-TiFPN

Design and Implementation of Hardware-in-the-loop Test Architecture for UAV Information Attack Based on QualNet

UAV Swarm On-line Mission Planning Method Based on Dynamic Allocation Strategy

Related Author

Hao TAN

Hang SU

Qifan TAN

Xuqi YANG

Xinyu CUI

Haotian LI

Mengzhen LIU

Guangyan HUANG

Related Institution

School of Future Aerospace Technology, Beihang University

Intelligent Collaborative Perception and Analytical Cognition Laboratory

School of Mechanical, Electronic and Control Engineering, Beijing Jiaotong University

School of Information and Electronics, Beijing Institute of Technology

Explosion Protection and Emergency Disposal Technology Engineering Research Center of the Ministry of Education, Beijing Institute of Technology

Postal code：100089
Tel：010-68963060/68962718 Fax：010-68963025 Email：bgxb@cos.org.cn
Technical support is provided by Beijing Founder electronics co., LTD 京ICP备05059581号-4 京公网安备11010802024360号
It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
Total Visits：208993 Visits Today：1

⁰