A High Accuracy and Large-Scale Detection for Fixed-Wing UAV Autonomous Ground Landing With GNSS-Denied

This paper concentrates on the ground-guided autonomous landing of a fixed-wing Unmanned Aerial Vehicle (UAV) within Global Navigation Satellite System (GNSS) denied environments. A highly accurate and large-scale automatic landing system for fixed-wing UAVs in GNSS-denied environments is proposed, leveraging advanced multi-sensor fusion techniques. To address the challenge of low accuracy over extended ranges, a distance measurement module is introduced as a key component. Additionally, the integration of near-infrared and visible cameras ensures precise UAV detection, thereby significantly enhancing the system’s reliability and overall performance. To further facilitate large-scale range detection in GNSS-denied conditions, a landing guidance window is developed, enabling the efficient search for and tracking of UAVs. Moreover, a joint calibration method (JCM) is implemented to minimize discrepancies between the system’s outputs and GNSS data. This method involves a systematic two-step calibration process: first, the GNSS and pan-tilt unit (PTU) coordinate systems are aligned on the runway, and then the PTU and landing point coordinate systems are calibrated. At the same time, the near-infrared and visible cameras are calibrated simultaneously to ensure synchronized and accurate performance across the system. To demonstrate the generalizability and robustness of this approach, the proposed system is implemented on two UAV platforms tested across diverse and complex scenarios. When compared to state-of-the-art methods, the proposed system not only achieves superior accuracy but also exhibits exceptional scalability, making it particularly well-suited for GNSS-denied environments.