MF-UAVPose6D: A Model-Free Monocular 6-DoF Pose Estimation Framework for Fixed-Wing UAVs

2026-06-28

AI Summary

This work addresses the challenge of 6-DoF pose estimation for non-cooperative fixed-wing UAVs, where the absence of CAD models and keypoint priors renders conventional methods ineffective. To this end, we propose the first monocular, model-free 6-DoF pose estimation framework that operates solely on a single RGB image and known camera intrinsics. Our approach leverages heatmap-guided center localization, a pose-aware module (PAM), dynamic topology sampling (DTS), and a decoupled translation-rotation decoding mechanism to effectively exploit structural cues. We also introduce FW-UAV6DPose, the first synthetic dataset encompassing diverse distances and viewpoints. Experimental results demonstrate that our method achieves robust performance in long-range rotation estimation, depth recovery, and joint pose evaluation, enabling accurate and efficient model-free pose estimation.

Abstract

For uncrewed aerial vehicles (UAVs), estimating six-degree-of-freedom (6-DoF) poses is essential for airspace situational awareness, target tracking, and counter-UAV operations. However, non-cooperative targets usually lack computer-aided design (CAD) models and keypoint priors, making existing model-based or keypoint-matching methods difficult to apply reliably. To address these challenges, this paper proposes MF-UAVPose6D, a model-free monocular 6-DoF pose estimation framework for fixed-wing UAVs. During inference, the method takes only a single red-green-blue (RGB) image and camera intrinsics as input. It first obtains a stable target anchor through heatmap-guided center localization, introduces a Perspective-Aware Module (PAM) to model observation-ray priors, exploits Dynamic Topological Sampling (DTS) to complement weak structural cues from the wings, fuselage, and tail, and adopts a decoupled translation-rotation pose decoding mechanism to estimate the 6-DoF pose. In addition, we construct the FW-UAV6DPose synthetic dataset, which covers fixed-wing UAV observations across diverse distances, viewpoints, and poses. Experimental results show that MF-UAVPose6D achieves accurate and efficient monocular 6-DoF pose estimation without requiring CAD models, and demonstrates strong robustness in long-range rotation estimation, depth recovery, and joint pose evaluation.

Problem

• 6-DoF pose estimation
• model-free
• fixed-wing UAVs
• monocular vision
• non-cooperative targets

Innovation

• model-free
• monocular 6-DoF pose estimation
• Perspective-Aware Module
• Dynamic Topological Sampling
• fixed-wing UAV