Vision-Based Risk Aware Emergency Landing for UAVs in Complex Urban Environments

Diego Mercado-Ravell; Israel Becerra; Julio de la Torre-Vanegas; Miguel Soriano-Garcia

arxiv: 2505.20423 · v1 · submitted 2025-05-26 · 💻 cs.RO · cs.AI· cs.CV

Vision-Based Risk Aware Emergency Landing for UAVs in Complex Urban Environments

Julio de la Torre-Vanegas , Miguel Soriano-Garcia , Israel Becerra , Diego Mercado-Ravell This is my paper

Pith reviewed 2026-05-19 12:32 UTC · model grok-4.3

classification 💻 cs.RO cs.AIcs.CV

keywords UAV emergency landingrisk-aware navigationsemantic segmentationsafe landing zoneurban environmentsdeep neural networkvision-based control

0 comments

The pith

A specialized neural network creates risk maps that let UAVs select safe landing zones in dynamic urban settings.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a vision-based system for UAV emergency landings in cities. It uses a deep neural network to assign risk values to each pixel in the camera view, identifying hazards like moving vehicles and people. An algorithm then selects a stable safe landing zone using these risk maps, and a control system directs the drone there with safety thresholds that change with altitude. This matters because current UAVs struggle with safe emergency descents in populated areas, where crashes could harm people or property. The approach was tested in real urban environments with over 90 percent success.

Core claim

The central claim is that by combining semantic segmentation for pixel-level risk assessment with risk-map algorithms and altitude-dependent control, UAVs can adaptively find and descend to safe landing zones in complex urban environments despite moving obstacles and illumination changes, achieving over 90% success rates in real-world tests.

What carries the argument

Pixel-level risk assignment via deep neural network feeding into risk maps for Safe Landing Zone (SLZ) identification and temporal stabilization in the descent controller.

Load-bearing premise

The specialized deep neural network reliably assigns accurate pixel-level risk values to moving obstacles and under shifting illumination without requiring extensive retraining or additional sensors.

What would settle it

Flight tests in busy urban areas with moving vehicles and pedestrians under varying light that yield landing success rates below 80 percent would falsify the central effectiveness claim.

Figures

Figures reproduced from arXiv: 2505.20423 by Diego Mercado-Ravell, Israel Becerra, Julio de la Torre-Vanegas, Miguel Soriano-Garcia.

**Figure 1.** Figure 1: Left: Drone-view image Ik from the UAV. Right: Grayscale representation of the risk map Rk, where brighter pixels denote higher risk. module MS. The module MS assigns each pixel to a category relevant to urban environments, such as grass, pavement, cars, people, vegetation, etc. Later, a risk mapping MR is applied such that each category is assigned a numerical risk value, with low values indicating safer … view at source ↗

**Figure 2.** Figure 2: Two consecutive image frames demonstrating the [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗

**Figure 3.** Figure 3: Progressive visualization of the risk expansion process. Top row shows 2D [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗

**Figure 4.** Figure 4: Visualization of the landing point selection process. (a) Filtered risk map [PITH_FULL_IMAGE:figures/full_fig_p012_4.png] view at source ↗

**Figure 5.** Figure 5: Visualization of the landing point stabilization process at different altitudes. [PITH_FULL_IMAGE:figures/full_fig_p013_5.png] view at source ↗

read the original abstract

Landing safely in crowded urban environments remains an essential yet challenging endeavor for Unmanned Aerial Vehicles (UAVs), especially in emergency situations. In this work, we propose a risk-aware approach that harnesses semantic segmentation to continuously evaluate potential hazards in the drone's field of view. By using a specialized deep neural network to assign pixel-level risk values and applying an algorithm based on risk maps, our method adaptively identifies a stable Safe Landing Zone (SLZ) despite moving critical obstacles such as vehicles, people, etc., and other visual challenges like shifting illumination. A control system then guides the UAV toward this low-risk region, employing altitude-dependent safety thresholds and temporal landing point stabilization to ensure robust descent trajectories. Experimental validation in diverse urban environments demonstrates the effectiveness of our approach, achieving over 90% landing success rates in very challenging real scenarios, showing significant improvements in various risk metrics. Our findings suggest that risk-oriented vision methods can effectively help reduce the risk of accidents in emergency landing situations, particularly in complex, unstructured, urban scenarios, densely populated with moving risky obstacles, while potentiating the true capabilities of UAVs in complex urban operations.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper puts together a workable vision pipeline for urban UAV emergency landing but the 90% success claim sits on unexamined assumptions about the DNN's handling of motion and light.

read the letter

The core contribution is a straightforward system that runs semantic segmentation to build pixel risk maps, picks a stable safe landing zone, and feeds an altitude-aware controller with temporal smoothing. That combination addresses a concrete operational need for drones in cities where sudden failures require quick descent without hitting people or vehicles. The temporal stabilization and altitude thresholds are practical touches that could reduce jitter in real flights, and the focus on moving obstacles shows awareness of urban realities rather than just static scene assumptions. Credit for shipping an integrated approach instead of stopping at segmentation alone. The experiments are the weak part. The abstract states over 90% success and risk improvements in diverse real urban settings, yet gives no trial counts, baseline comparisons, per-class accuracy on dynamic objects, or illumination tests. Without those, the central claim that the risk DNN generalizes reliably stays unverified. The stress-test concern lands: if the network was not explicitly checked on moving targets under changing light, both the reported gains and the safety guarantees rest on hope rather than data. This is the sort of work that belongs in a robotics applications venue or a safety-focused conference. Engineers building delivery or inspection drones would find the pipeline ideas useful as a starting point, even if they end up redoing the validation. A serious editor should send it to review once the authors add the missing quantitative details and ablations; the idea is grounded enough to merit referee time.

Referee Report

2 major / 2 minor

Summary. The manuscript proposes a vision-based risk-aware emergency landing system for UAVs operating in complex urban environments. It uses semantic segmentation via a specialized deep neural network to produce pixel-level risk maps, applies a risk-map algorithm to identify a temporally stabilized Safe Landing Zone (SLZ) while handling moving obstacles such as vehicles and pedestrians, and employs an altitude-dependent control law with safety thresholds to guide descent. The central empirical claim is that the integrated pipeline achieves over 90% landing success rates in challenging real-world urban scenarios and yields measurable improvements on various risk metrics.

Significance. If the experimental claims are substantiated with adequate quantitative detail, the work would offer a practical contribution to UAV autonomy by demonstrating a sensor-light, vision-only method for dynamic risk assessment and safe emergency landing in unstructured urban settings. The combination of semantic risk mapping with altitude-aware control and temporal stabilization addresses a relevant operational gap in robotics applications for drones.

major comments (2)

[Experimental Results] Experimental Results section: The claim of >90% landing success and risk-metric improvements is presented without reporting the total number of trials, the number or diversity of urban test environments, the specific baseline methods compared against, or any statistical measures (error bars, confidence intervals, or p-values). This absence prevents assessment of whether the reported performance is robust or sensitive to post-hoc experimental choices.
[Risk Assignment Network] Risk Assignment Network subsection: The pipeline's reliability hinges on the DNN producing accurate pixel-level risk values for moving obstacles and under illumination shifts without retraining or extra sensors. No per-class IoU or precision-recall figures for dynamic object categories, no illumination-ablation results, and no cross-environment transfer metrics are supplied to support this generalization claim.

minor comments (2)

[Abstract] Abstract: The phrase 'significant improvements in various risk metrics' is used without naming the metrics or quantifying the gains; adding one concrete example would improve clarity.
[Figures and Notation] Notation and figures: Ensure that all acronyms (SLZ, DNN) are defined on first use and that figure captions explicitly state the units or scale of any risk-map visualizations.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive feedback on our manuscript. The comments highlight important aspects for strengthening the experimental validation and the supporting evidence for the risk assignment network's reliability. We address each major comment below and will make the necessary revisions to the manuscript.

read point-by-point responses

Referee: [Experimental Results] Experimental Results section: The claim of >90% landing success and risk-metric improvements is presented without reporting the total number of trials, the number or diversity of urban test environments, the specific baseline methods compared against, or any statistical measures (error bars, confidence intervals, or p-values). This absence prevents assessment of whether the reported performance is robust or sensitive to post-hoc experimental choices.

Authors: We agree that the Experimental Results section lacks sufficient detail to allow a full assessment of the robustness of our claims. In the revised manuscript, we will provide the total number of trials conducted, describe the number and diversity of urban test environments used, specify the baseline methods against which comparisons were made, and include statistical measures such as error bars, confidence intervals, and where appropriate p-values. These additions will clarify the experimental setup and support the reported performance metrics. revision: yes
Referee: [Risk Assignment Network] Risk Assignment Network subsection: The pipeline's reliability hinges on the DNN producing accurate pixel-level risk values for moving obstacles and under illumination shifts without retraining or extra sensors. No per-class IoU or precision-recall figures for dynamic object categories, no illumination-ablation results, and no cross-environment transfer metrics are supplied to support this generalization claim.

Authors: We recognize the need for more quantitative support regarding the generalization of the Risk Assignment Network. Although the manuscript describes the network's design for handling moving obstacles and illumination changes, we will supplement the Risk Assignment Network subsection with per-class IoU and precision-recall figures for dynamic categories, results from illumination ablation studies, and cross-environment transfer metrics. This will provide stronger evidence for the network's performance without requiring retraining or additional sensors. revision: yes

Circularity Check

0 steps flagged

No circularity in empirical pipeline

full rationale

The paper presents an applied robotics system that combines semantic segmentation via a specialized DNN to produce pixel-level risk maps, an algorithm to select stable safe landing zones (SLZs) while handling moving obstacles and illumination changes, and an altitude-dependent controller with temporal stabilization. Performance claims rest on reported experimental success rates (>90%) from real-world urban tests rather than any closed-form derivation, uniqueness theorem, or fitted parameter that is then relabeled as a prediction. No equations, self-definitional steps, or load-bearing self-citations appear in the abstract or description that would reduce the central results to their own inputs by construction; the work is therefore self-contained as an empirical demonstration.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract provides no explicit free parameters, axioms, or invented entities; the method implicitly assumes a pre-trained segmentation network and risk-threshold rules whose details are not stated.

pith-pipeline@v0.9.0 · 5748 in / 1042 out tokens · 32799 ms · 2026-05-19T12:32:55.214535+00:00 · methodology

discussion (0)

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

See&Say: Vision Language Guided Safe Zone Detection for Autonomous Package Delivery Drones
cs.CV 2026-04 unverdicted novelty 5.0

See&Say combines depth gradients, semantic masks, and VLM-guided refinement to generate safety maps and alternative drop zones for autonomous drone deliveries, outperforming baselines in accuracy and IoU.

Reference graph

Works this paper leans on

28 extracted references · 28 canonical work pages · cited by 1 Pith paper

[1]

IEEE Access12(2024)

H. Shakhatreh, A. H. Sawalmeh, A. Al-Fuqaha, Z. Dou, E. Almaita, I. Khalil, N. S. Othman, A. Khreishah, M. Guizani, Unmanned aerial vehicles (uavs): A survey on civil applications and key research chal- lenges, IEEE Access 7 (2019) 48572–48634.doi:10.1109/ACCESS. 2019.2909530. 22

work page doi:10.1109/access 2019
[2]

A. V. R. Katkuri, H. Madan, N. Khatri, A. S. H. Abdul-Qawy, K. S. Patnaik, Autonomous UAV navigation using deep learning-based com- putervisionframeworks: Asystematicliteraturereview, Array23(2024) 100361.doi:10.1016/j.array.2024.100361

work page doi:10.1016/j.array.2024.100361 2024
[3]

Shah Alam, J

M. Shah Alam, J. Oluoch, A survey of safe landing zone de- tection techniques for autonomous unmanned aerial vehicles (uavs), Expert Systems with Applications 179 (2021) 115091. doi:https://doi.org/10.1016/j.eswa.2021.115091. URLhttps://www.sciencedirect.com/science/article/pii/ S0957417421005327

work page doi:10.1016/j.eswa.2021.115091 2021
[4]

Pieczyński, B

D. Pieczyński, B. Ptak, M. Kraft, M. Piechocki, P. Aszkowski, A fast, lightweight deep learning vision pipeline for autonomous UAV land- ing support with added robustness, Engineering Applications of Arti- ficial Intelligence 131 (2024) 107864.doi:10.1016/j.engappai.2024. 107864

work page doi:10.1016/j.engappai.2024 2024
[5]

Demirhan, C

M. Demirhan, C. Premachandra, Development of an automated camera- based drone landing system, IEEE Access 8 (2020) 202111–202121.doi: 10.1109/ACCESS.2020.3034948

work page doi:10.1109/access.2020.3034948 2020
[6]

M. Lv, B. Fan, J. Fang, J. Wang, Autonomous landing of quadrotor unmanned aerial vehicles based on multi-level marker and linear active disturbance reject control, Sensors 24 (5) (2024) 1645

work page 2024
[7]

A. E. Morando, M. F. Santos, P. Castillo, A. Correa-Victorino, Vision- Based Algorithm for Autonomous Aerial Landing, in: 2024 International Conference on Unmanned Aircraft Systems (ICUAS), IEEE, Chania - Crete, Greece, 2024, pp. 652–657.doi:10.1109/ICUAS60882.2024. 10556880

work page doi:10.1109/icuas60882.2024 2024
[8]

H. Xu, L. Wang, W. Han, Y. Yang, J. Li, Y. Lu, J. Li, A sur- vey on uav applications in smart city management: Challenges, ad- vances, and opportunities, IEEE Journal of Selected Topics in Ap- plied Earth Observations and Remote Sensing 16 (2023) 8982–9010. doi:10.1109/JSTARS.2023.3317500

work page doi:10.1109/jstars.2023.3317500 2023
[9]

joint authorities for rulemaking of unmanned systems (2024)

Jarus guidelines on specific operations risk assessment (sora) v2.0. joint authorities for rulemaking of unmanned systems (2024). 23

work page 2024
[10]

J. A. Loera-Ponce, D. A. Mercado-Ravell, I. Becerra, L. M. Valentin- Coronado, Risk assessment for uav autonomous landing in urban en- vironments using semantic segmentation, in: Advances in Artificial In- telligence – IBERAMIA 2024, Springer Nature Switzerland, 2025, pp. 197–208

work page 2024
[11]

Abdollahzadeh, P.-L

S. Abdollahzadeh, P.-L. Proulx, M. S. Allili, J.-F. Lapointe, Safe land- ing zones detection for uavs using deep regression, in: 2022 19th Conference on Robots and Vision (CRV), 2022, pp. 213–218.doi: 10.1109/CRV55824.2022.00035

work page doi:10.1109/crv55824.2022.00035 2022
[12]

Benjwal, P

A. Benjwal, P. Uday, A. Vadduri, A. Pai, Safe landing zone detec- tion for uavs using image segmentation and super resolution, in: In- ternational Conference on Machine Vision Applications, 2023.doi: 10.23919/MVA57639.2023.10215759

work page doi:10.23919/mva57639.2023.10215759 2023
[13]

Soriano-García, D

M. Soriano-García, D. Mercado-Ravell, Viva-safeland: a new freeware for safe validation of vision-based navigation in aerial vehicles (2025). arXiv:2503.14719. URLhttps://arxiv.org/abs/2503.14719

work page arXiv 2025
[14]

T. Yang, P. Li, H. Zhang, J. Li, Z. Li, Monocular vision slam-based uav autonomous landing in emergencies and unknown environments, Electronics 7 (5) (2018). URLhttps://www.mdpi.com/2079-9292/7/5/73

work page 2018
[15]

M. A. Kaljahi, P. Shivakumara, M. Y. I. Idris, M. H. Anisi, T. Lu, M. Blumenstein, N. M. Noor, An automatic zone detection system for safe landing of uavs, Expert Systems with Applications 122 (2019) 319–333.doi:https://doi.org/10.1016/j.eswa.2019.01.024. URLhttps://www.sciencedirect.com/science/article/pii/ S0957417419300272

work page doi:10.1016/j.eswa.2019.01.024 2019
[16]

Saldiran, M

E. Saldiran, M. Hasanzade, A. Cetin, G. Inalhan, Autonomous emer- gency landing system to unknown terrain for uavs, in: 2024 AIAA DATC/IEEE 43rd Digital Avionics Systems Conference (DASC), 2024, pp. 1–7.doi:10.1109/DASC62030.2024.10749002

work page doi:10.1109/dasc62030.2024.10749002 2024
[17]

Saldiran, M

E. Saldiran, M. Hasanzade, A. Cetin, G. Inalhan, Ensuring operation time safety of vtol uav: Autonomous emergency landings in unknown 24 terrain, IEEE Transactions on Aerospace and Electronic Systems (2025). doi:10.1109/TAES.2025.3559899

work page doi:10.1109/taes.2025.3559899 2025
[18]

L. Xin, Z. Tang, W. Gai, H. Liu, Vision-based autonomous landing for the uav: A review, Aerospace 9 (11) (2022). URLhttps://www.mdpi.com/2226-4310/9/11/634

work page 2022
[19]

Zhang, Q

Q. Zhang, Q. Xia, L. Wei, B. Deng, A vision-based method for uav au- tonomous landing area detection, in: Z. Shi, M. Witbrock, Q. Tian (Eds.), Intelligence Science V, Springer Nature Switzerland, Cham, 2025, pp. 204–213

work page 2025
[20]

J. Wu, Z. Zhang, W. Huang, Semantic map construction of uav au- tonomous landing in unknown environment, in: 2024 36th Chinese Control and Decision Conference (CCDC), 2024, pp. 5018–5025.doi: 10.1109/CCDC62350.2024.10588362

work page doi:10.1109/ccdc62350.2024.10588362 2024
[21]

Secchiero, N

M. Secchiero, N. Bobbili, Y. Zhou, G. Loianno, Visual environment as- sessment for safe autonomous quadrotor landing, in: 2024 International ConferenceonUnmannedAircraftSystems(ICUAS),2024, pp.807–813. doi:10.1109/ICUAS60882.2024.10557078

work page doi:10.1109/icuas60882.2024.10557078 2024
[22]

Tzelepi, A

M. Tzelepi, A. Tefas, Graph embedded convolutional neural networks in human crowd detection for drone flight safety, IEEE Transactions on Emerging Topics in Computational Intelligence 5 (2) (2021) 191–204. doi:10.1109/TETCI.2019.2897815

work page doi:10.1109/tetci.2019.2897815 2021
[23]

González-Trejo, D

J. González-Trejo, D. Mercado-Ravell, I. Becerra, R. Murrieta-Cid, On thevisual-basedsafelandingofuavsinpopulatedareas: Acrucialaspect for urban deployment, IEEE Robotics and Automation Letters 6 (4) (2021) 7901–7908

work page 2021
[24]

J. Lim, M. Kim, H. Yoo, J. Lee, Autonomous multirotor uav search and landing on safe spots based on combined semantic and depth informa- tion from an onboard camera and lidar, IEEE/ASME Transactions on Mechatronics 29 (5) (2024) 3960–3970

work page 2024
[25]

Mitroudas, V

T. Mitroudas, V. Balaska, A. Psomoulis, A. Gasteratos, Light-weight approach for safe landing in populated areas, IEEE International Con- ference on Robotics and Automation (2024) 10027–10032. 25

work page 2024
[26]

N. S. Vemulapalli, P. Paladugula, G. S. Prabhat, S. Don, Reinforce- ment Learning-Based Autonomous Landing of AirSim Simulated Quad- copter in Unreal Engine, in: 2024 15th International Conference on Computing Communication and Networking Technologies (ICCCNT), IEEE, Kamand, India, 2024, pp. 1–7.doi:10.1109/ICCCNT61001. 2024.10725648

work page doi:10.1109/icccnt61001 2024
[27]

S. Liu, W. Li, H. Li, S. Li, Reinforcement learning based multi- perspective motion planning of manned electric vertical take-off and landing vehicle in urban environment with wind fields, Engineering Applications of Artificial Intelligence 149 (2025) 110392.doi:https: //doi.org/10.1016/j.engappai.2025.110392

work page doi:10.1016/j.engappai.2025.110392 2025
[28]

Tovanche-Picon, J

H. Tovanche-Picon, J. González-Trejo, Á. Flores-Abad, M. Á. García- Terán, D. Mercado-Ravell, Real-time safe validation of autonomous landing in populated areas: From virtual environments to Robot- In-The-Loop, Virtual Reality 28 (1) (2024) 66.doi:10.1007/ s10055-024-00965-6. 26

work page 2024

[1] [1]

IEEE Access12(2024)

H. Shakhatreh, A. H. Sawalmeh, A. Al-Fuqaha, Z. Dou, E. Almaita, I. Khalil, N. S. Othman, A. Khreishah, M. Guizani, Unmanned aerial vehicles (uavs): A survey on civil applications and key research chal- lenges, IEEE Access 7 (2019) 48572–48634.doi:10.1109/ACCESS. 2019.2909530. 22

work page doi:10.1109/access 2019

[2] [2]

A. V. R. Katkuri, H. Madan, N. Khatri, A. S. H. Abdul-Qawy, K. S. Patnaik, Autonomous UAV navigation using deep learning-based com- putervisionframeworks: Asystematicliteraturereview, Array23(2024) 100361.doi:10.1016/j.array.2024.100361

work page doi:10.1016/j.array.2024.100361 2024

[3] [3]

Shah Alam, J

M. Shah Alam, J. Oluoch, A survey of safe landing zone de- tection techniques for autonomous unmanned aerial vehicles (uavs), Expert Systems with Applications 179 (2021) 115091. doi:https://doi.org/10.1016/j.eswa.2021.115091. URLhttps://www.sciencedirect.com/science/article/pii/ S0957417421005327

work page doi:10.1016/j.eswa.2021.115091 2021

[4] [4]

Pieczyński, B

D. Pieczyński, B. Ptak, M. Kraft, M. Piechocki, P. Aszkowski, A fast, lightweight deep learning vision pipeline for autonomous UAV land- ing support with added robustness, Engineering Applications of Arti- ficial Intelligence 131 (2024) 107864.doi:10.1016/j.engappai.2024. 107864

work page doi:10.1016/j.engappai.2024 2024

[5] [5]

Demirhan, C

M. Demirhan, C. Premachandra, Development of an automated camera- based drone landing system, IEEE Access 8 (2020) 202111–202121.doi: 10.1109/ACCESS.2020.3034948

work page doi:10.1109/access.2020.3034948 2020

[6] [6]

M. Lv, B. Fan, J. Fang, J. Wang, Autonomous landing of quadrotor unmanned aerial vehicles based on multi-level marker and linear active disturbance reject control, Sensors 24 (5) (2024) 1645

work page 2024

[7] [7]

A. E. Morando, M. F. Santos, P. Castillo, A. Correa-Victorino, Vision- Based Algorithm for Autonomous Aerial Landing, in: 2024 International Conference on Unmanned Aircraft Systems (ICUAS), IEEE, Chania - Crete, Greece, 2024, pp. 652–657.doi:10.1109/ICUAS60882.2024. 10556880

work page doi:10.1109/icuas60882.2024 2024

[8] [8]

H. Xu, L. Wang, W. Han, Y. Yang, J. Li, Y. Lu, J. Li, A sur- vey on uav applications in smart city management: Challenges, ad- vances, and opportunities, IEEE Journal of Selected Topics in Ap- plied Earth Observations and Remote Sensing 16 (2023) 8982–9010. doi:10.1109/JSTARS.2023.3317500

work page doi:10.1109/jstars.2023.3317500 2023

[9] [9]

joint authorities for rulemaking of unmanned systems (2024)

Jarus guidelines on specific operations risk assessment (sora) v2.0. joint authorities for rulemaking of unmanned systems (2024). 23

work page 2024

[10] [10]

J. A. Loera-Ponce, D. A. Mercado-Ravell, I. Becerra, L. M. Valentin- Coronado, Risk assessment for uav autonomous landing in urban en- vironments using semantic segmentation, in: Advances in Artificial In- telligence – IBERAMIA 2024, Springer Nature Switzerland, 2025, pp. 197–208

work page 2024

[11] [11]

Abdollahzadeh, P.-L

S. Abdollahzadeh, P.-L. Proulx, M. S. Allili, J.-F. Lapointe, Safe land- ing zones detection for uavs using deep regression, in: 2022 19th Conference on Robots and Vision (CRV), 2022, pp. 213–218.doi: 10.1109/CRV55824.2022.00035

work page doi:10.1109/crv55824.2022.00035 2022

[12] [12]

Benjwal, P

A. Benjwal, P. Uday, A. Vadduri, A. Pai, Safe landing zone detec- tion for uavs using image segmentation and super resolution, in: In- ternational Conference on Machine Vision Applications, 2023.doi: 10.23919/MVA57639.2023.10215759

work page doi:10.23919/mva57639.2023.10215759 2023

[13] [13]

Soriano-García, D

M. Soriano-García, D. Mercado-Ravell, Viva-safeland: a new freeware for safe validation of vision-based navigation in aerial vehicles (2025). arXiv:2503.14719. URLhttps://arxiv.org/abs/2503.14719

work page arXiv 2025

[14] [14]

T. Yang, P. Li, H. Zhang, J. Li, Z. Li, Monocular vision slam-based uav autonomous landing in emergencies and unknown environments, Electronics 7 (5) (2018). URLhttps://www.mdpi.com/2079-9292/7/5/73

work page 2018

[15] [15]

M. A. Kaljahi, P. Shivakumara, M. Y. I. Idris, M. H. Anisi, T. Lu, M. Blumenstein, N. M. Noor, An automatic zone detection system for safe landing of uavs, Expert Systems with Applications 122 (2019) 319–333.doi:https://doi.org/10.1016/j.eswa.2019.01.024. URLhttps://www.sciencedirect.com/science/article/pii/ S0957417419300272

work page doi:10.1016/j.eswa.2019.01.024 2019

[16] [16]

Saldiran, M

E. Saldiran, M. Hasanzade, A. Cetin, G. Inalhan, Autonomous emer- gency landing system to unknown terrain for uavs, in: 2024 AIAA DATC/IEEE 43rd Digital Avionics Systems Conference (DASC), 2024, pp. 1–7.doi:10.1109/DASC62030.2024.10749002

work page doi:10.1109/dasc62030.2024.10749002 2024

[17] [17]

Saldiran, M

E. Saldiran, M. Hasanzade, A. Cetin, G. Inalhan, Ensuring operation time safety of vtol uav: Autonomous emergency landings in unknown 24 terrain, IEEE Transactions on Aerospace and Electronic Systems (2025). doi:10.1109/TAES.2025.3559899

work page doi:10.1109/taes.2025.3559899 2025

[18] [18]

L. Xin, Z. Tang, W. Gai, H. Liu, Vision-based autonomous landing for the uav: A review, Aerospace 9 (11) (2022). URLhttps://www.mdpi.com/2226-4310/9/11/634

work page 2022

[19] [19]

Zhang, Q

Q. Zhang, Q. Xia, L. Wei, B. Deng, A vision-based method for uav au- tonomous landing area detection, in: Z. Shi, M. Witbrock, Q. Tian (Eds.), Intelligence Science V, Springer Nature Switzerland, Cham, 2025, pp. 204–213

work page 2025

[20] [20]

J. Wu, Z. Zhang, W. Huang, Semantic map construction of uav au- tonomous landing in unknown environment, in: 2024 36th Chinese Control and Decision Conference (CCDC), 2024, pp. 5018–5025.doi: 10.1109/CCDC62350.2024.10588362

work page doi:10.1109/ccdc62350.2024.10588362 2024

[21] [21]

Secchiero, N

M. Secchiero, N. Bobbili, Y. Zhou, G. Loianno, Visual environment as- sessment for safe autonomous quadrotor landing, in: 2024 International ConferenceonUnmannedAircraftSystems(ICUAS),2024, pp.807–813. doi:10.1109/ICUAS60882.2024.10557078

work page doi:10.1109/icuas60882.2024.10557078 2024

[22] [22]

Tzelepi, A

M. Tzelepi, A. Tefas, Graph embedded convolutional neural networks in human crowd detection for drone flight safety, IEEE Transactions on Emerging Topics in Computational Intelligence 5 (2) (2021) 191–204. doi:10.1109/TETCI.2019.2897815

work page doi:10.1109/tetci.2019.2897815 2021

[23] [23]

González-Trejo, D

J. González-Trejo, D. Mercado-Ravell, I. Becerra, R. Murrieta-Cid, On thevisual-basedsafelandingofuavsinpopulatedareas: Acrucialaspect for urban deployment, IEEE Robotics and Automation Letters 6 (4) (2021) 7901–7908

work page 2021

[24] [24]

J. Lim, M. Kim, H. Yoo, J. Lee, Autonomous multirotor uav search and landing on safe spots based on combined semantic and depth informa- tion from an onboard camera and lidar, IEEE/ASME Transactions on Mechatronics 29 (5) (2024) 3960–3970

work page 2024

[25] [25]

Mitroudas, V

T. Mitroudas, V. Balaska, A. Psomoulis, A. Gasteratos, Light-weight approach for safe landing in populated areas, IEEE International Con- ference on Robotics and Automation (2024) 10027–10032. 25

work page 2024

[26] [26]

N. S. Vemulapalli, P. Paladugula, G. S. Prabhat, S. Don, Reinforce- ment Learning-Based Autonomous Landing of AirSim Simulated Quad- copter in Unreal Engine, in: 2024 15th International Conference on Computing Communication and Networking Technologies (ICCCNT), IEEE, Kamand, India, 2024, pp. 1–7.doi:10.1109/ICCCNT61001. 2024.10725648

work page doi:10.1109/icccnt61001 2024

[27] [27]

S. Liu, W. Li, H. Li, S. Li, Reinforcement learning based multi- perspective motion planning of manned electric vertical take-off and landing vehicle in urban environment with wind fields, Engineering Applications of Artificial Intelligence 149 (2025) 110392.doi:https: //doi.org/10.1016/j.engappai.2025.110392

work page doi:10.1016/j.engappai.2025.110392 2025

[28] [28]

Tovanche-Picon, J

H. Tovanche-Picon, J. González-Trejo, Á. Flores-Abad, M. Á. García- Terán, D. Mercado-Ravell, Real-time safe validation of autonomous landing in populated areas: From virtual environments to Robot- In-The-Loop, Virtual Reality 28 (1) (2024) 66.doi:10.1007/ s10055-024-00965-6. 26

work page 2024