A signal-processing approach to terrain following for remote-sensing remotely piloted aerial systems




Skelly, Robert Mark William

Journal Title

Journal ISSN

Volume Title



Recent advances in technology have enabled an explosion of interest in compact remotely piloted aerial systems (RPAS). RPAS are of particular interest to remote sensing scientists, as they provide an accessible, low-cost way to perform surveys over small sites at extremely high levels of detail. The interaction between low flight altitude and high relative terrain relief introduces safety, efficiency and data-quality issues which can only be resolved by closely following the terrain. A terrain following system would ensure that specific measures of data quality -- point density, the signal-to-noise ratio and scale distortion -- are held as near as possible to constant. Safety and efficiency would be addressed by ensuring that a terrain following trajectory does not contain gradients or inflections that exceed the vehicle's dynamic limits, in particular its maximum acceleration and velocity. This research develops a forward-looking, predictive terrain following system using a signal-processing approach, in which the terrain itself is conceived as a signal and filtered appropriately to extract a trajectory which achieves these objectives.



remote sensing, remotely piloted aerial system, rpas, unmanned aerial vehicle, uav, lidar, hyperspectral, savitzky-golay, signal processing