Monocular visual SLAM for small UAVs in GPS-denied environments

This paper presents a monocular visual simultaneous localization and mapping (SLAM) system for a small unmanned aerial vehicle (UAV) in GPS-denied environments. A single camera is used to measure the motion of the vehicle and detect features (landmarks) for the map building. The SLAM estimates the positions of the UAV and the features by an extended Kalman filter (EKF), which takes the velocity estimated by the fusion of the inertial and visual measurements as the input. An inverse depth method is adopted for the feature initialization. Both simulations and experiments are carried out to verify the effectiveness of this system.