VisLanding: Monocular 3D Perception for UAV Safe Landing via Depth-Normal Synergy

2025-06-17

AI Summary

To address the challenge of safe, autonomous UAV landing in complex and unknown environments, this paper proposes a monocular 3D perception framework that formulates Safe Landing Zone (SLZ) estimation as an end-to-end binary semantic segmentation task. Methodologically, it introduces a novel joint regression mechanism for depth and surface normal vectors, integrates a fine-tuned Metric3D V2 backbone, and designs a dedicated SLZ segmentation head. We further construct WildUAV, a new dataset featuring drone-view annotations and cross-domain evaluation subsets. Key contributions include: (1) the first monocular-image-based framework enabling joint quantitative estimation of SLZ location and area; (2) zero-shot cross-domain generalization capability, significantly improving robustness and segmentation accuracy; and (3) real-time performance and operational reliability validated on an actual UAV decision-making system.

Technology Category

• Computer Vision: 3D Computer Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Machine Learning: Calibration & Uncertainty Quantification

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Security and Privacy: Large-scale security measurements
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

This paper presents VisLanding, a monocular 3D perception-based framework for safe UAV (Unmanned Aerial Vehicle) landing. Addressing the core challenge of autonomous UAV landing in complex and unknown environments, this study innovatively leverages the depth-normal synergy prediction capabilities of the Metric3D V2 model to construct an end-to-end safe landing zones (SLZ) estimation framework. By introducing a safe zone segmentation branch, we transform the landing zone estimation task into a binary semantic segmentation problem. The model is fine-tuned and annotated using the WildUAV dataset from a UAV perspective, while a cross-domain evaluation dataset is constructed to validate the model's robustness. Experimental results demonstrate that VisLanding significantly enhances the accuracy of safe zone identification through a depth-normal joint optimization mechanism, while retaining the zero-shot generalization advantages of Metric3D V2. The proposed method exhibits superior generalization and robustness in cross-domain testing compared to other approaches. Furthermore, it enables the estimation of landing zone area by integrating predicted depth and normal information, providing critical decision-making support for practical applications.

Problem

• Monocular 3D perception for UAV safe landing
• Depth-normal synergy for safe zone identification
• Cross-domain robustness in complex environments

Innovation

• Monocular 3D perception for UAV landing
• Depth-normal synergy prediction via Metric3D V2
• Safe zone segmentation as binary classification