基于低秩特征增强的飞行器景象匹配定位方法

王永海; 何琪彬; 郭灵犀; 陈超; 薛晗庆

doi:10.7527/S1000-6893.2026.33468

航空学报 >

0 1 - 0

DOI: https://doi.org/10.7527/S1000-6893.2026.33468

基于低秩特征增强的飞行器景象匹配定位方法

王永海 ,
何琪彬 ,
郭灵犀 ,
陈超 ,
薛晗庆

展开

1. 中国运载火箭技术研究院
2. 临近空间物理重点实验室
3. 空间物理重点实验室

收稿日期: 2026-02-04

修回日期: 2026-03-31

网络出版日期: 2026-04-02

收起

Low-rank feature-enhanced scene matching for aircraft localization

WANG Yong-Hai ,
HE Qi-Bin ,
GUO Ling-Xi ,
CHEN Chao ,
XUE Han-Qing

Expand

Received date: 2026-02-04

Revised date: 2026-03-31

Online published: 2026-04-02

Fold

摘要

景象匹配是解决卫星导航拒止环境下飞行器自主定位问题的关键技术,对于提升飞行器在视觉结构丰富区域的定位能力、支撑其在临近空间等高动态环境中的可靠应用具有重要价值。现有基于深度学习的方法难以有效区分图像中稳定本征结构与瞬时干扰噪声,导致其在面对剧烈视角、季节、模态等复杂域变化时泛化能力不足,且缺乏明确物理先验引导以保障鲁棒性。为此,本文提出了一种低秩特征增强的景象匹配方法。通过将低秩先验嵌入深度神经网络,构建了端到端的低秩特征增强网络(Low-rank Feature Enhancement Network,LFE-Net)框架,利用Schatten-p 范数损失隐式约束模型聚焦于场景稳定结构,并结合多任务学习提升泛化性能。实验表明,该方法在飞行器景象匹配数据集上取得了更高的平均定位精度,对复杂域变化表现出强鲁棒性。

关键词： 景象匹配; 视觉地理定位; 端到端学习; 低秩特征增强

本文引用格式

王永海 , 何琪彬 , 郭灵犀 , 陈超 , 薛晗庆 . 基于低秩特征增强的飞行器景象匹配定位方法[J]. 航空学报, 0 : 1 -0 . DOI: 10.7527/S1000-6893.2026.33468

Abstract

Scene matching is a key technology for solving the problem of autonomous positioning of aircraft in satellite navigation denied environments. It is of great value for improving the positioning capability of aircraft in visually rich regions and supporting their reliable application in highly dynamic environments such as near space. Existing deep learning-based methods struggle to effectively distinguish stable intrinsic structures from transient noise in images, resulting in insufficient generalization ability when faced with complex domain changes, e.g., drastic changes in viewpoint, season, and modality. Furthermore, they lack explicit physical prior guidance to ensure robustness. As a consequence, this paper proposes a low-rank feature enhancement-based scene matching method. By embedding low-rank priors into a deep neural network, an end-to-end Low-rank Feature Enhancement Network (LFE-Net) framework is constructed. The Schatten-p norm loss implicit constraint model focuses on stable scene structures, and multi-task learning is combined to improve generalization performance. Experiments show that this method achieves higher average localization accuracy on aircraft scene matching datasets and exhibits strong robustness to complex domain changes.

Key words： scene matching; visual geo-location; end-to-end learning; low-rank feature enhancement

参考文献

[1] 赵春晖, 刘安萌, 吕洋, 潘泉. 无人机韧性自主定位技术综述[J]. 航空学报, 2024, 45(8): 28839. ZHAO C, LIU A, LYU Y, PAN Q. A survey of resilient selflocalization for UAV[J]. Acta Aeronautica et Astronautica Sinica, 2024, 45(8): 28839 (in Chinese).
[2] Zhu J, Yan S, Wang L, et al. LoD-Loc: Aerial visual localization using LoD 3D map with neural wireframe alignment[C]//Advances in Neural Information Processing Systems. 2024, 37.
[3] 唐彬, 杨小冈, 卢瑞涛, 张震宇, 宿爽. 基于图像翻译的飞行器红外/卫星异源快速匹配定位方法 [J]. 航空学报, 2025, 46(23): 631961. TANG B, YANG X, LU R, ZHANG Z, SU S. Aircraft infrared/satellite heterogenous fast matching localization method based on image translation[J]. Acta Aeronautica et Astronautica Sinica, 2025, 46(23): 631961.
[4] 蓝朝桢, 阎晓东, 崔志祥, 等. 用于无人机自主绝对定位的实时特征匹配方法 [J]. 测绘科学, 2023, 48(3): 264-268. LAN C Z, YAN X D, CUI Z X, et al. Real-time feature matching method for the autonomous absolute location of UAV[J]. Science of Surveying and Mapping, 2023, 48(3): 264-268 (in Chinese).
[5] 袁媛, 孙柏, 刘赶超. 景象匹配无人机视觉定位 [J]. 自动化学报, 2024, 51: 1-25. YUAN Y, SUN B, LIU G C. Visual Positioning of UAVs via Scene Matching[J]. Acta Automatica Sinica, 2024, 51: 1-25 (in Chinese).
[6] Zheng Z, Wei Y, Yang Y. University-1652: A multiview multi-source benchmark for drone-based geolocalization[C]//Proceedings of the 28th ACM international conference on Multimedia. 2020: 1395-1403.
[7] Lu W N, et al. Semi-distributed cross-modal air-ground relative localization for limited bandwidth conditions[C]//IEEE/RSJ International Conference on Intelligent Robots and Systems. 2025: 1-8.
[8] 沈林成; 卜彦龙; 徐昕; 潘亮. 景象匹配辅助组合导航中景象区域适配性研究进展 [J]. 航空学报, 2010, 31(3): 553-563. Shen L;Bu Y;Xu X;Pan L. Research on Matching-area Suitability for Scene Matching Aided Navigation[J]. ACTA AERONAUTICAET ASTRONAUTICA SINICA, 2010, 31(3): 553-563 (in Chinese).
[9] Lowe D G. Distinctive image features from scale-invariant keypoints[J]. International journal of computer vision, 2004, 60(2): 91-110.
[10] 辛瑞, 张霄力, 彭侠夫, 等. 结合区域全局特征和 Ann-SIFT 的二阶段快速景象匹配算法 [J]. 电子技术应用, 2023, 49(5): 135-141. XIN R, ZHANG X L, PENG X F, et al. Two-stage fast scene matching algorithm combining global descriptor and AnnSIFT[J]. Application of Electronic Technique, 2023, 49(5): 135141 (in Chinese).
[11] Bay H, Ess A, Tuytelaars T, et al. Speeded-up robust features (SURF)[J]. Computer vision and image understanding, 2008, 110(3): 346-359.
[12] Zhu R, Yin L, Yang M, et al. SUES-200: A multi-height multiscene cross-view image benchmark across drone and satellite[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2023, 33(9): 4825-4839.
[13] 牛燕雄, 陈梦琪, 张贺. 基于尺度不变特征变换的快速景象匹配方法 [J]. 电子与信息学报, 2019, 41(3): 626-631. NIU Y X, CHEN M Q, ZHANG H. Fast scene matching method based on scale invariant feature transform[J]. Journal of Electronics & Information Technology, 2019, 41(3): 626-631 (in Chinese).
[14] Wang Y, Feng X, Li F, et al. Lightweight visual localization algorithm for UAVs[J]. Scientific Reports, 2025, 15(1): 6069.
[15] 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.
[16] Xia Z, Booij O, Manfredi M, et al. Visual cross-view metric localization with dense uncertainty estimates[C]//European Conference on Computer Vision. Cham: Springer Nature Switzerland, 2022: 90-106.
[17] 王华夏, 程咏梅, 刘楠. 面向山地区域光照变化下的鲁棒景象匹配方法 [J]. 航空学报, 2017, 38(10): 321101-321101. WANG H, CHENG Y, LIU N. A robust scene matching method for mountainous regions with illumination variation[J]. ACTA AERONAUTICAET ASTRONAUTICA SINICA, 2017, 38(10): 321101-321101 (in Chinese).
[18] He Q, Yan Z, Diao W, et al. Dlc: dynamic loss correction for crossdomain remotely sensed segmentation[J]. IEEE Transactions on Geoscience and Remote Sensing, 2024, 62: 1-14.
[19] Yu Z, Zhu B, Han Z, et al. Reliability analysis of multimodal scene matching based on correlation peaks[J]. Chinese Journal of Aeronautics, 2024, 37(9): 433-447.
[20] 屈若锟, 王致远, 刘晔璐, 李诚龙, 江波. 面向城市空中交通的无人机视觉定位技术 [J]. 航空学报, 2025, 46(11): 531168. QU R, WANG Z, LIU Y, LI C, JIANG B. UAV visual positioning technology for urban air mobility[J]. Acta Aeronautica et Astronautica Sinica, 2025, 46(11): 531168 (in Chinese).
[21] Dai M, Zheng E, Chen J, et al. Drone Referring Localization: An Efficient Heterogeneous Spatial Feature Interaction Method For UAV Self-Localization[J]. arXiv preprint arXiv:2408.06561, 2024.
[22] 杨鸿睿, 朱启举, 曹培贤, 等. 一种用于景象匹配导航的新型图像配准算法 [J]. 工程科学学报, 2025, 47(3): 496-503. YANG H R, ZHU Q J, CAO P X, et al. A novel image registration algorithm for scene matching navigation[J]. Chinese Journal of Engineering, 2025, 47(3): 496-503 (in Chinese).
[23] Zhang X, Qin H, Ma L, et al. Deep Feature Matching of Differentmodal Images for Visual Geo-Localization of UAVs[J]. IEEE Transactions on Aerospace and Electronic Systems, 2024.
[24] Xia P, Wan Y, Zheng Z, et al. Enhancing cross-view geolocalization with domain alignment and scene consistency[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2024.
[25] Zeng Q, Wu J, Ren Y, et al. Cross-view geolocation via segmentation and common region feature matching[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2025.
[26] Liu T, Ren K, Chen Q. Towards natural language-guided drones: GeoText-1652 benchmark with spatial relation matching[C]//European Conference on Computer Vision. 2024: 112128.
[27] 聂伟, 戴琪霏, 杨小龙, 等. 基于多维信号特征的无人机探测识别方法 [J]. 电子与信息学报, 2024, 46(3): 1089-1099. NIE W, DAI Q F, YANG X L, et al. UAV detection and recognition method based on multi-dimensional signal characteristics[J]. Journal of Electronics & Information Technology, 2024, 46(3): 1089-1099 (in Chinese).
[28] Ji Y, He B, Tan Z, et al. Game4loc: A uav geo-localization benchmark from game data[C]//Proceedings of the AAAI Conference on Artificial Intelligence. 2025, 39(4): 3913-3921.
[29] Liu X, Wang Z, Wu Y, et al. Segcn: A semantic-aware graph convolutional network for uav geo-localization[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2024, 17: 6055-6066.
[30] He Q, Xiao Z, Huang Z, et al. Orientation-aware multi-modal learning for road intersection identification and mapping[C]//2024 IEEE International Conference on Robotics and Automation. IEEE, 2024: 16185-16191.
[31] Schroff F, Kalenichenko D, Philbin J. Facenet: A unified embedding for face recognition and clustering[C]//Proceedings of the IEEE conference on computer vision and pattern recognition. 2015: 815-823.
[32] Ding L, Zhou J, Meng L, et al. A practical cross-view image matching method between UAV and satellite for UAV-based geolocalization[J]. Remote Sensing, 2020, 13(1): 47.
[33] Wang T, Zheng Z, Yan C, et al. Each part matters: Local patterns facilitate cross-view geo-localization[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2021, 32(2): 867879.
[34] Lin J, Zheng Z, Zhong Z, et al. Joint representation learning and keypoint detection for cross-view geo-localization[J]. IEEE Transactions on Image Processing, 2022, 31: 3780-3792.
[35] Deng Y, Lin X, Li R, et al. Multi-scale gem pooling with N-pair center loss for fine-grained image search[C]//2019 IEEE International Conference on Multimedia and Expo (ICME). IEEE, 2019: 1000-1005.
[36] Shen T, Wei Y, Kang L, et al. MCCG: A ConvNeXt-based multiple-classifier method for cross-view geo-localization[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2023, 34(3): 1456-1468.
[37] Dai M, Hu J, Zhuang J, et al. A transformer-based feature segmentation and region alignment method for UAV-view geolocalization[J]. IEEE Transactions on Circuits and Systems for Video Technology, 2022, 32(7): 4376-4389.
[38] He K, Zhang X, Ren S, et al. Deep residual learning for image recognition[C]//Proceedings of the IEEE conference on computer vision and pattern recognition. 2016: 770-778.
[39] Liu Z, Mao H, Wu C Y, et al. A convnet for the 2020s[C]//Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2022: 11976-11986.
[40] Wu H, Xiao B, Codella N, et al. Cvt: Introducing convolutions to vision transformers[C]//Proceedings of the IEEE/CVF international conference on computer vision. 2021: 22-31.
[41] Paszke A, Gross S, Massa F, et al. Pytorch: An imperative style, high-performance deep learning library[J]. Advances in neural information processing systems, 2019, 32.
[42] Loshchilov I, Hutter F. Decoupled weight decay regularization[J]. arXiv preprint arXiv:1711.05101, 2017.
[43] Dosovitskiy A, Beyer L, Kolesnikov A, et al. An image is worth 16x16 words: Transformers for image recognition at scale[C]//International Conference on Learning Representations, 2021: 1395-1403.

Options

文章导航

模态框（Modal）标题

摘要

本文引用格式

Abstract

参考文献