Camera View Selection for Bearing-Only Geolocation of Stationary Targets Using Multi-UAV Swarms with GPS Bias

Abstract

We address the problem of geolocating a static ground target from multiple low-cost Unmanned Aerial Vehicles (UAVs) equipped with monocular cameras and object detectors. Each drone generates large sequences of bearing-only measurements per flyover, but these are affected by per-drone GPS biases (systematic position errors specific to each UAV) that remain effectively constant over short trajectories. We propose a greedy selector that combines angular diversity and detection confidence to pre-filter measurements before Random Sample Consensus (RANSAC)-based robust fusion. Across 30 randomized simulation runs, the selector achieves mean accuracy comparable to full-pool RANSAC (2.34 m vs. 2.44 m) while reducing median processing time by 4.1× (393 ms vs. 1602 ms). Similar accuracy is observed against LO-RANSAC and PROSAC, with the selector remaining 4× to 35× faster. The selector also outperforms confidence-only and standard Fisher Information Matrix (FIM)-based subset baselines. A bias-aware FIM analysis via Schur-complement marginalization shows that the information contributed by repeated measurements from a single drone saturates under a shared position bias, explaining the single-drone concentration observed for these baselines. In the nominal setting, the angular-diversity proxy retains 99.7% of the bias-aware FIM objective and achieves localization error similar to direct FIM-based selection, while being 3.3× faster at the selection stage. Real flights with a low-cost quadrotor measure short-term GPS drift of 1.9–3.5 m over 10–63 min, supporting the assumed bias dynamics, and expose a failure mode in which persistent false positives capture the RANSAC consensus. Full-pool RANSAC shows slightly lower mean error at larger GPS-bias levels ( σ_GPS ≥ 5 m), indicating that pre-selection is not preferable in every operating regime.

Keywords

UAV geolocation / Bearing-only localization / Fisher Information Matrix / View selection / RANSAC / Drone swarms / Search and rescue (SAR)

References

[1] Sambolek S, Ivasic—Kos M. Person Detection and Geolocation Estimation in Drone Images. SN Comput. Sci. 2025, 6, 358. DOI: 10.1007/s42979—025—03869—7

[2] Jocher G, Chaurasia A, Qiu J. Ultralytics YOLOv8; Ultralytics, 2023. Available online: https://github.com/ultralytics/ultralytics (accessed on 19 May 2026).

[3] Lampesberger H, Hermann E, Zeilinger M. Object Geolocalization Using Consumer—Grade Devices: Algorithms and Applications. In Proceedings of the 4th International Workshop on Camera Traps, AI, and Ecology (CamTraps 2024), Hagenberg, Austria, 5—6 September 2024; pp. 71-83. DOI: 10.5281/zenodo.13740788

[4] Doğançay K. Bearings—only target localization using total least squares. Signal Process. 2005, 85, 1695-1710. DOI: 10.1016/j.sigpro.2005.03.007

[5] Ranacher P, Brunauer R, Trutschnig W, Van der Spek S, Reich S. Why GPS makes distances bigger than they are. Int. J. Geogr. Inf. Sci. 2016, 30, 316-333. DOI: 10.1080/13658816.2015.1086924

[6] Opromolla R. Magnetometer Calibration for Small Unmanned Aerial Vehicles Using Cooperative Flight Data. Sensors 2020, 20, 538. DOI: 10.3390/s20020538

[7] Bishop AN, Fidan B, Anderson BDO, Doğançay K, Pathirana PN. Optimality analysis of sensor—target localization geometries. Automatica 2010, 46, 479-492. DOI: 10.1016/j.automatica.2009.12.003

[8] Shamaiah M, Banerjee S, Vikalo H. Greedy Sensor Selection: Leveraging Submodularity. In Proceedings of the IEEE Conference on Decision and Control (CDC), Atlanta, GA, USA, 15—17 December 2010; pp. 2572—2577. DOI: 10.1109/CDC.2010.5717225

[9] Kirchner MR, Hespanha JP, Garagić D. Heterogeneous Measurement Selection for Vehicle Tracking using Submodular Optimization. In Proceedings of the 2020 IEEE Aerospace Conference, Big Sky, MT, USA, 7—14 March 2020; pp. 1-10. DOI: 10.1109/AERO47225.2020.9172788

[10] Xiong K, Wu X, Duan A, Huang Y, Leng S, He J. Information—Optimal Formation Geometry Design for Multimodal UAV Cooperative Perception. arXiv 2025, arXiv:2512.12923. DOI: 10.48550/arXiv.2512.12923

[11] Lee CW, Waslander SL. UncertaintyTrack: Exploiting Detection and Localization Uncertainty in Multi—Object Tracking. In Proceedings of 2024 IEEE International Conference on Robotics and Automation (ICRA), Yokohama, Japan, 13—17 May 2024; pp. 4946-4953. DOI: 10.1109/ICRA57147.2024.10610458

[12] Solano—Carrillo E, Sattler F, Alex A, Klein A, Costa BP, Rodriguez AB, et al. UTrack: Multi—object Tracking with Uncertain Detections. In Proceedings of the ECCV Workshops, Milan, Italy, 29 September—4 October 2024. DOI: 10.1007/978—3—031—91585—7_14

[13] Jung H, Kang S, Kim T, Kim H. ConfTrack: Kalman Filter—Based Multi—Person Tracking by Utilizing Confidence Score of Detection Box. In Proceedings of the 2024 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), Waikoloa, HI, USA, 3—8 January 2024; pp. 6569-6578. DOI: 10.1109/WACV57701.2024.00645

[14] Paulin G, Sambolek S, Ivasic—Kos M. Application of raycast method for person geolocalization and distance determination using UAV images in Real—World land search and rescue scenarios. Expert Syst. Appl. 2024, 237, 121495. DOI: 10.1016/j.eswa.2023.121495

[15] Sciacchitano A, Mudrik L, Kragelund S, Kaminer I. Multi—UAV Trajectory Optimization for Bearing—Only Localization in GPS Denied Environments. arXiv 2026, arXiv:2602.11116. DOI: 10.48550/arXiv.2602.11116

[16] Kaplan LM. Global node selection for localization in a distributed sensor network. IEEE Trans. Aerosp. Electron. Syst. 2006, 42, 113-135. DOI: 10.1109/TAES.2006.1603409

[17] Dogancay K. 3D Pseudolinear Target Motion Analysis From Angle Measurements. IEEE Trans. Signal Process. 2015, 63, 1570-1580. DOI: 10.1109/TSP.2015.2399869

[18] Nemhauser GL, Wolsey LA, Fisher ML. An analysis of approximations for maximizing submodular set functions—I. Math. Program. 1978, 14, 265-294. DOI: 10.1007/BF01588971

[19] Joshi S, Boyd S. Sensor Selection via Convex Optimization. IEEE Trans. Signal Process. 2009, 57, 451-462. DOI: 10.1109/TSP.2008.2007095

[20] Krause A, Singh A, Guestrin C. Near—Optimal Sensor Placements in Gaussian Processes: Theory, Efficient Algorithms and Empirical Studies. J. Mach. Learn. Res. 2008, 9, 235-284. Available online: https://www.jmlr.org/papers/volume9/krause08a.html (accessed on 19 May 2026).

[21] Lin B, Wu L, Niu Y. End—to—End Vision—Based Cooperative Target Geo—Localization for Multiple Micro UAVs. J. Intell. Robot. Syst. 2022, 106, 13. DOI: 10.1007/s10846—022—01639—8

[22] Sun P, Tong S, Qin K, Luo Z, Lin B, Shi M. Low—Cost Real—Time Remote Sensing and Geolocation of Moving Targets via Monocular Bearing—Only Micro UAVs. Remote Sens. 2025, 17, 3836. DOI: 10.3390/rs17233836

[23] Georgiadis D, Politi E, Solmaz S, Dimitrakopoulos G. A review of localization and sensing technologies for UAV swarms in SAR missions. EURASIP J. Adv. Signal Process. 2026, 2026, 8. DOI: 10.1186/s13634—025—01269—w

[24] Fischler MA, Bolles RC. Random sample consensus. Commun. ACM 1981, 24, 381-395. DOI: 10.1145/358669.358692

[25] Chum O, Matas J. Matching with PROSAC—Progressive Sample Consensus. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), San Diego, CA, USA, 20—25 June 2005; pp. 220-226. DOI: 10.1109/CVPR.2005.221

[26] Chum O, Matas J, Kittler J. Locally Optimized RANSAC. In Proceedings of the DAGM Symposium on Pattern Recognition, Magdeburg, Germany, 10—12 September 2003; pp. 236-243. DOI: 10.1007/978—3—540—45243—0_31

[27] Raguram R, Chum O, Pollefeys M, Matas J, Frahm JM. USAC: A Universal Framework for Random Sample Consensus. IEEE Trans. Pattern Anal. Mach. Intell. 2013, 35, 2022-2038. DOI: 10.1109/TPAMI.2012.257

[28] Hartley R, Zisserman A. Multiple View Geometry in Computer Vision, 2nd ed.; Cambridge University Press: Cambridge, UK, 2004. DOI: 10.1017/CBO9780511811685

[29] Jiang B, Luo R, Mao J, Xiao T, Jiang Y. Acquisition of Localization Confidence for Accurate Object Detection. In Proceedings of the European Conference on Computer Vision (ECCV), Munich, Germany, 8—14 September 2018. DOI: 10.1007/978—3—030—01264—9_48

[30] Zhu P, Wen L, Du D, Bian X, Fan H, Hu Q, et al. Detection and Tracking Meet Drones Challenge. IEEE Trans. Pattern Anal. Mach. Intell. 2021, 44, 7380-7399. DOI: 10.1109/TPAMI.2021.3119563

[31] PUC—Rio GrADyS Project. GrADyS—Sim NextGen. Available online: https://github.com/Project—GrADyS/gradys—sim—nextgen (accessed on 19 May 2026).