Position Based Visual Servoing with Artificial Neural Networks for Quadrotor-type Unmanned Aerial Vehicles

Abstract

UAVs offer many advantages over manned vehicles and their application area expands in time passes. Particularly, interest in UAVs with rotary wing types is increasing. Increasing interest in these vehicles has spawned many different controller designs. In this study, estimating the pose of the vehicle according to the image features with the help of a single camera mounted on the quadrotor and then position-based visual servoing, a technique that allows the vehicle to be controlled by using the image features, was used. In this study position-based visual servoing (PBVS); 3D parameter estimates of the vehicle pose were implemented with artificial neural networks.