Automation of take-off and landing relies on accurate knowledge of variables, such as the position, velocity and orientation of the vehicle. While GPS is a popular sensor choice in open spaces, it suffers from accuracy issues in urban environments. We therefore use cameras to estimate both the state of the vehicle and its surrounding obstacles. The demo showed a Micro Aerial Vehicle (MAV) that solely uses a camera system for stabilization and navigation. It can be steered by an untrained pilot with a joystick, while it avoids crashing into obstacles.
Watch a video
here.