Air-Ground Collaborative Landing in Dynamic Lighting Environments via Onboard Infrared-Visible Feedback Fusion

This paper presents a distributed air-ground system that achieves collaborative landing in dynamic lighting environments via infrared-RGB sensor fusion. By integrating RGB and IR cameras via a weighted fusion algorithm, the proposed system leverages complementary visual and thermal information to robustly enhance target recognition under sudden illumination changes. An active perception strategy incorporating observation-aware trajectory planning allows the unmanned aerial vehicle (UAV) to dynamically adjust its landing path based on real-time target detection. Experimental results under three lighting scenarios, including sudden drops in illumination, demonstrate that the fused dual-modal system consistently identifies the target and enables accurate landing, whereas traditional RGB-only recognition fails. The results verify the robustness and adaptability of the proposed approach, providing a practical solution for reliable UAV cooperative landing in complex lighting conditions.