Real-time 6DoF deck pose estimation and target tracking for landing an UAV in a cluttered shipboard environment using on-board vision

Abstract

We present a novel vision system for estimating the 6DoF relative pose of an UAV from a shipdeck. The vision system aims to detect the international landing target using the edge information of the target to perform line segment mapping, feature point clustering and filtering. A novel target-shape evaluation scheme is developed to fuse the measurements. Real-time onboard flight tests demonstrate that the improved vision system is fast, accurate and capable of dealing with an incomplete target, whilst the overall design is able to perform target tracking and achieve closed-loop control.