Use este identificador para citar ou linkar para este item:
Título: 3D forest structure analysis from optical and LIDAR data
Autor(es): Robert Treuhaft
Bruce Chapman
Luciano Dutra
João Roberto dos Santos
Fábio Gonçalves
José Claúdio Mura
Paulo Maurício Lima de Alencastro Graça
Jason Drake
Assunto: forest profile
radar interferometry
ISSN: 1808-0251
Revista: Ambiência (UNICENTRO)
Volume: 2
Resumo: In Austria about half of the entire area (46 %) is covered by forests. The majority of these are highly managed and controlled in growth. But besides timber production forest ecosystems play a multifunctional role including climate control, habitat provision and, especially in Austria, protection of settlements. The interrelationships among climatic, ecological, social and economic dimensions of forests require technologies for monitoring both the state and the development of forests. This comprises forest structure, species and age composition and, forest integrity in general. Assessing forest structure for example enables forest managers and natural risk engineers to evaluate whether a forest can fulfil its protective function or not. Traditional methods for assessing forest structure like field inventories and aerial photo interpretation are intrinsically limited in providing spatially continuous information over a large area. The Centre for Geoinformatics (Z_GIS) in collaboration with the National Park Bayerischer Wald, Germany and the Stand Montafon, Austria, has tested and applied advanced approaches of integrating multispectral optical data and airborne laser scanning (ALS) data for (1) forest stand delineation, (2) single tree detection and (3) forest structure analysis. As optical data we used RGBI line scanner data and CIR air-photos. ALS data were raw point data (10 pulses per sqm) and normalised crown models (nCM) at 0.5 m and 1 m resolution. (1) Automated stand delineation was done by (a) translating a key for manual mapping of forest development phases into a rule-based system via object-relationship modelling (ORM); and (b) by performing multiresolution segmentation and GIS analysis. (2) Strategies forsingle tree detection using raw ALS data included (a) GIS modelling based on aregion-growth local maxima algorith mand (b) object-based image analysis using super class information class-specific rule sets. (3) Verticalforest structure has been assessed statistically by (a) applying basic statistics (like mean, standard deviation, and variation coefficient) on the raw data using a moving window approach; and (b) by applying landscape metrics (Shannon Evenness Index, SHEI, and division ndex, DIVI) for different strata extracted from the nCM.
ISSN: 1808-0251
Aparece nas coleções:Coordenação de Dinâmica Ambiental (CDAM)

Arquivos associados a este item:
Arquivo Descrição TamanhoFormato 
tres_D_forest.pdf511,19 kBAdobe PDFVisualizar/Abrir

Este item está licenciada sob uma Licença Creative Commons Creative Commons