Vision Based Autonomous Perching Approach

Vision-based autonomous perching for aerial robots aims to enable drones to land and securely attach themselves to various surfaces, conserving energy and extending mission duration. Current research emphasizes robust methods for detecting and approaching perching targets, often employing computer vision techniques with multiple markers for improved accuracy and reliability, coupled with control algorithms (including reinforcement learning and trajectory optimization) to manage the landing maneuver, even on moving or inclined surfaces. This technology has significant implications for applications such as environmental monitoring, infrastructure inspection, and search and rescue, where extended operational time and stable observation points are crucial.

Papers