The work presents the comparison of results obtained with two different methods for filtering LiDAR data, focusing specifically on ground (bare Earth) and vegetation classification. Data was filtered with TerraScan, a leading proprietary software for processing LiDAR point clouds, and with the specific modules available in GRASS GIS. The LiDAR dataset (1.5 points/m 2 ) was produced with an Optech ALTM Gemini and covers an area of the Sardinia region where there are available also orthophotos and false color infrared images, both with resolution of 20 cm. Concerning TerraScan two classifications were computed: the former is a completely automatic and the latter is a semi-automatic one, with a heavy manual reclassification of the previously obtained automatic results. In GRASS GIS the modules developed by the Geomatic Laboratory of the Politecnico di Milano were applied; parameters to be used by the modules were at first calibrated on specific training sub-areas. The purpose of the work is the evaluation of the accuracy in ground and vegetation extraction. Moreover, since GRASS divides points belonging to vegetation in two classes, corresponding to high and low vegetation, the comparison is done for the latter category. The choice of concentrating on this category is motivated by the greater difficulty in distinguishing between terrain and low vegetation. As “ground truth” the result of the supervised classification performed on the orthophotos and the false color infrared images was used.

Comparison of GRASS-LiDAR modules–TerraScan with respect to vegetation filtering

BROVELLI, MARIA ANTONIA;LUCCA, SARA
2012-01-01

Abstract

The work presents the comparison of results obtained with two different methods for filtering LiDAR data, focusing specifically on ground (bare Earth) and vegetation classification. Data was filtered with TerraScan, a leading proprietary software for processing LiDAR point clouds, and with the specific modules available in GRASS GIS. The LiDAR dataset (1.5 points/m 2 ) was produced with an Optech ALTM Gemini and covers an area of the Sardinia region where there are available also orthophotos and false color infrared images, both with resolution of 20 cm. Concerning TerraScan two classifications were computed: the former is a completely automatic and the latter is a semi-automatic one, with a heavy manual reclassification of the previously obtained automatic results. In GRASS GIS the modules developed by the Geomatic Laboratory of the Politecnico di Milano were applied; parameters to be used by the modules were at first calibrated on specific training sub-areas. The purpose of the work is the evaluation of the accuracy in ground and vegetation extraction. Moreover, since GRASS divides points belonging to vegetation in two classes, corresponding to high and low vegetation, the comparison is done for the latter category. The choice of concentrating on this category is motivated by the greater difficulty in distinguishing between terrain and low vegetation. As “ground truth” the result of the supervised classification performed on the orthophotos and the false color infrared images was used.
2012
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11311/661210
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 3
  • ???jsp.display-item.citation.isi??? ND
social impact