|Synergy of airborne digital camera and Lidar data to map coastal dune vegetation|Kempeneers, P.; Deronde, B.; Provoost, S.; Houthuys, R. (2009). Synergy of airborne digital camera and Lidar data to map coastal dune vegetation. J. Coast. Res. SI 53: 73-82. dx.doi.org/10.2112/SI53-009.1
In: Journal of Coastal Research. Coastal Education and Research Foundation: Fort Lauderdale. ISSN 0749-0208, more
lidar; vegetation classification; mapping; digital elevation models;
Driven by the successful applications of lidar in forestry and the availability of lidar technology, new research is being carried out in other ecosystems. While lidar data have often been used to study tall forest ecosystems, this Study explores the utility of lidar in the lower-canopy ecosystems of the Belgian coastal dune belt. This area is largely covered by marram dune, moss dune, grassland, scrubs and some woodland. Small diameter (0.4 m) footprint lidar was applied to derive the canopy height by analyzing the first and last pulse returns simultaneously. The investigation focused on whether the height of low-canopy ecosystems Could be mapped with adequate accuracy. An error analysis was performed first on flat terrain (i.e., tennis court and parking lot) and then on vegetation canopy. The mapping of coastal dune vegetation is necessary to establish the strength of the dune belt. Dune vegetation fixes the sand dunes, protecting them from erosion and from possible breakthroughs threatening the historically reclaimed land (polders) situated inland from the dunes. Next, multispectral data was acquired from a digital camera with Visual and near infrared channels. The digital camera overflight was not conducted on the same platform as the lidar. After ortho-rectification of the multispectral image, the data of both sources were fused. The limited spectral information delivered by the digital camera was not able to provide a sufficiently detailed and accurate vegetation map. The fusion with lidar data provided the extra information needed to obtain the desired vegetation and dune strength maps. A total of fourteen classes were defined, of which twelve cover vegetation. It was shown that overall classification accuracy improved 16%, from 55% to 71% after data fusion.