Vision-Aided Navigation for Small UAVs in GPS-Challenged Environments

Abstract

As UAVs become smaller, their flight space will expand to include flying within urban canyons and eventually inside buildings. Flying within these environments is very challenging from a navigation perspective, as GPS is largely denied or degraded by dropouts and multi-path eects. Traditional navigation approaches based on filtering GPS and lowcost inertial sensors produce unacceptably high drift during extended GPS outages. A promising remedy is to augment navigation with an onboard vision sensor that tracks visual landmarks to infer vehicle motion. We propose a vision-aided navigation system that can replace GPS when it is temporarily denied. The system uses a single camera as the vision sensor, because many UAVs already carry a camera as part of their limited payload. The system detects, tracks, and geo-locates 3D landmarks observed in the stream of camera images. We consider several options for initializing the range to landmarks in the navigation filter, including motion stereo. The system geo-locates a set of landmarks while GPS provides accurate navigation. Once GPS is lost, the system combines estimated landmark locations with new observations of these landmarks to both navigate and geolocate new landmarks in a process called boot-strapping. Tests of the system in simulated and on flight data collected with a small UAV show that vision-aided navigation drift is significantly lower than under inertial-only navigation.