Indoor UAV Localization and 3D Mapping Using Visual Odometry

Outdoor operations of autonomous UAVs (Un- manned Aerial Vehicle)highly rely on GPS sensors. Whereat localization is essential for UAVs (Unmanned Aerial Vehicle) to operate properly. Positioning by GPS isn’t feasible in indoor environments due to signal attenuation. In this study a flight controller which uses GPS for localization has been made operational in indoor environments by using reverse engineering approach and visual odometry data. Visual odometry was produced by Intel® RealSenseTM T265 and RTAB-Map were run via depth image which was acquired from Intel® RealSenseTM D435i on UAV.