Reinforcement Learning-Based Monocular Vision Approach for Autonomous UAV Landing
This addresses the problem of cost-effective and efficient autonomous landing for UAVs, offering a potential alternative to complex sensor setups, though it appears incremental in leveraging reinforcement learning for a specific visual task.
The paper tackles autonomous UAV landing using only a monocular camera by reframing it as an optimization problem based on visual cues from a landing pad, achieving robust and accurate results in simulations and experiments without depth sensors.
This paper introduces an innovative approach for the autonomous landing of Unmanned Aerial Vehicles (UAVs) using only a front-facing monocular camera, therefore obviating the requirement for depth estimation cameras. Drawing on the inherent human estimating process, the proposed method reframes the landing task as an optimization problem. The UAV employs variations in the visual characteristics of a specially designed lenticular circle on the landing pad, where the perceived color and form provide critical information for estimating both altitude and depth. Reinforcement learning algorithms are utilized to approximate the functions governing these estimations, enabling the UAV to ascertain ideal landing settings via training. This method's efficacy is assessed by simulations and experiments, showcasing its potential for robust and accurate autonomous landing without dependence on complex sensor setups. This research contributes to the advancement of cost-effective and efficient UAV landing solutions, paving the way for wider applicability across various fields.