Robustness Enhancement on Visually Guided UAV Autonomous Landing

Autonomous landing is a critical technology in drone applications, particularly in delivery and mission returns, ensuring a safe landing to successfully complete tasks. However, most existing research on autonomous landing remains limited to ideal environments, resulting in a significant gap between theoretical models and real-world applications. One of the primary challenges is enabling drones to autonomously perceive their surroundings and adjust their landing behavior accordingly. Mainstream autonomous landing technologies integrate visual tracking, trajectory prediction, and path planning to handle dynamic environments and complex challenges. This study addresses issues such as high-altitude operations, wind field variations, and occluded targets by designing a vision-based autonomous landing system. By combining visual algorithms and path planning techniques, this system improves the resilience and adaptability of drone landings in dynamic wind conditions, thus broadening its practical application value.