[Back]

N. Pfeifer, G. Mandlburger:
"Filtering and DTM Generation";
in: "Topographic Laser Ranging and Scanning: Principles and Processing", J. Shan, C. Toth (ed.); CRC Press, 2008, (invited), ISBN: 9781420051421, 307 - 333.

In this chapter, digital elevation models (DEM), also known as digital terrain models (DTM), will be discussed. After defining these terms we will first consider the specific advantages LiDAR offers for capturing the terrain surface, also in comparison to image-
based photogrammetric techniques. The main topic in this chapter will be, however,
the extraction of the terrain surface from the LiDAR measurements, either the original point cloud or the already preprocessed one with terrain elevations sorted into a raster. This process is a classification task, dividing the points or pixels into either ground or off-
terrain, where some approaches derive more classes. The usage of full-waveform echo recording for terrain reconstruction will be also discussed within this scope. Computing
a DTM from the results is the next step, but general discussion on strict interpolation
and qualified approximation techniques will be very limited. However, in order to make a DTM or DEM usable for many disciplines, e.g., hydraulics, geology, etc., the so-
called structure or feature lines should also be integrated. Likewise, the amount of data is typically very high and not adapted to the variations in terrain roughness. We will therefore continue with a section on the determination and exploitation of structure
lines from the LiDAR data as well as with methods for reducing the data amount, i.e., compressing the elevation data, considering error bounds. The quality of DTMs from airborne LiDAR data will be treated in a separate section. Real-world examples, embedded
in the text, will demonstrate the above. While DTMs can also be reconstructed from terrestrial laser scanning data, offering more detailed and accurate description, this is only possible for small areas in the order of 1 km2 and below, whereas airborne laser scanning has been used to collect topographic information over more than 10,000 km2 in dedicated projects. Therefore, only airborne laser scanning, also termed airborne scanning
LiDAR, will be considered.