Feature-Based Perception-Aware Multi-UAV Trajectory Planning

This work presents a trajectory planning framework that enhances multi-agent UAV navigation in GPS-denied environments. By leveraging visual feature observations between agents, the method accounts for estimator uncertainty and maintains a coherent shared map through multi-agent frame alignment. A perception-aware reward further guides UAVs toward trajectories with strong feature visibility and cross-agent redundancy. Experiments with two UAVs confirm that alignment significantly reduces relative distance error, while simulations show improved coordination without sacrificing goal-reaching ability. Together, these results highlight a promising approach for more reliable multi-agent navigation in challenging environments.