 |
| Titel |
Integration of Aerial and long-range Terrestrial Laser Scanner: alignment strategies and 3D models for landslide/rockslide description |
| VerfasserIn |
Eleonora Bertacchini, Riccardo Rivola, Cristina Castagnetti, Alessandro Capra, Erika Parmeggiani |
| Konferenz |
EGU General Assembly 2013
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 15 (2013) |
| Datensatznummer |
250083498
|
|
|
|
|
|
| Zusammenfassung |
| The purpose of this work is to integrate airborne and long-range (up to some kilometers) terrestrial laser scanning data in order to obtain a good and reliable description of the geomorphology of unstable slopes. 3D models derived from the integrated point clouds can be particularly useful to study and analyze complex morphology, vertical walls of rockslides and surfaces with sparse vegetation.
Terrestrial Laser Scanning (TLS) and Airborne Laser Scanning (ALS) are very efficient techniques for characterizing the morphostructure of slopes; although both techniques individually show some limits, their integration can overcome these limits.
For example, for vertical walls of rock, the TLS could be recommended because of its “frontal” point of view, that permits a description of vertical walls with a higher point density than ALS can reach. On the other hand, the power of the ALS methodology is highlighted when a large and flat area is involved and/or when a lot of vegetation covers the area. The ALS can easily reach the ground with respect to TLS due to its vertical position of measurement and this allows a more reliable result even if the last generation of terrestrial laser scanners has the skill to penetrate inside the vegetation thanks to a multi-echo technology.
All the problems and difficulties encountered will be fully described. When a reliable description of the morphology is requested and involved distances are of some kilometers, the alignment strategies, the point cloud management and filtering, along with the DTM generation are crucial aspects that cannot be underestimate. Thus, different alignment techniques were examined (ICP - Iterative Closest Point, backsigthing orientation with GPS positioning, alignment with targets or homologous points). Results could be optimized only combining the different alignment approaches. In addition, the DTM generation was deeply analyzed, comparing TIN (Triangular Irregular Network) and grid mesh format, different point cloud densities, different grid dimensions and algorithms of interpolation. The challenge was the definition of a quantitative method for comparing 3D models obtained with the different approaches. |
|
|
|
|
|
|