The paper deals with monocular image-based sense and avoid assuming constant aircraft velocities and straight flight paths. From very limited two dimensional image information it finally characterizes the whole three dimensional collision situation by estimating the time to closest point of approach, the horizontal relative distance and its direction and the vertical relative distance also. The distances are relative to the intruder aircraft horizontal and vertical sizes. The overall estimated relative distance is the closest between the two aircraft in three dimension. So finally, every important information can be extracted to be used in a collision decision. The applicability of the developed method is presented in software-in-the-loop simulation test runs. Several intruder size and speed values are considered together with trajectories covering the whole three dimensional space. The horizontal intruder flight directions relative to the own aircraft cover Systems and Control Laboratory, MTA SZTAKI, H-1111 Budapest, Kende utca 13-17.
This paper proposes a vision-integrated navigation system to guide an aircraft on the final glide path. It makes use of onboard vision systems which track runway features and estimate a 6D aircraft pose with respect to a runway to land. The proposed vision-integrated navigation system will allow an aircraft to continue the final approach procedure by maintaining the navigation precision in case of possible degradation or failure of ILS or GNSS/SBAS sensors. In order to handle a non-negligible delay of such vision-based measurements due to the image processing time, an error-sate Kalman filter (ESKF) framework incooporating time-delayed measurements is established. The proposed delayed-measurement ESKF framework uses a fact that camera image acquisitions are triggered by a system and hence can be notified without delay. This enables the navigation filter to perform back-propagation of the estimated state forward in time to prepare for the future correction step at the time the measurement becomes available. The vision-integrated navigation system based on this framework was developed and its functionality is validated in simulations. Its estimation performance will be flight-evaluated with two different vision systems onboard a fixed-wing UAV experimental platform.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.