With the ability to collect full-waveform signals from airborne LiDAR systems researchers from the Vienna University of Technology wanted to see if the distribution of amplitude and echo width could be used to predict whether a point was terrain or “other”. This could lead to more automated data processing and increased DTM accuracy.
Their research examined different ground cover conditions to insure that their approach would not be biased. Using the full-waveform methodology they were able to accurately label 92% of the terrain echoes correctly. However, they concluded that amplitude and echo width were not sufficient to completely identify terrain points, but that this processing could be used as part of a larger classification methodology.
Automating LiDAR data processing is the key to opening up the use of this data.