Wyniki wyszukiwania - BazTech

Ograniczanie wyników

1 2005

Powiadomienia systemowe

Sesja wygasła!
Sesja wygasła!

Znaleziono wyników: 1

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: Delanuay algorithm

Sortuj według:

Ogranicz wyniki do:

Zastosowanie klasyfikacji obiektowej wysokorozdzielczych obrazów teledetekcyjnych oraz analiz przestrzennych GIS w kartowaniu drzewostanów

Kok R., Kozioł K., Wężyk P.

Roczniki Geomatyki

2005

T. 3, z. 4

99-108

Single crown detection in forest stands based on VHR imagery has been the subject of research for over a decade. Recent progress in crown detection shows how hotspots can be isolated from tree crowns. The spatial distribution of hotspots is related to tree species, stand development and clown density. Due to high resolution (spectral & spatial) of the QuickBird data set, insight into the relationship between crown-hotspots, tree age and crown density - can be analyzed in more detail. The data fusion of spatial information obtained by forest inventory, GPS measurements and the relation between SILP attribute and geometrical database (digital forest map . LMN) allows to build up the connection to the spectral signature registered by the satellite imagery. Estimation of crown diameter based on the hotspots from Quickbird data, as well as the spatial distribution of the total crown area is the main goal of this study. Various concepts (Thyssen polygon, irregular triangulation etc) open diverse modeling possibilities. In this study, a selection of available models is applied to the point set of crownhotspots. For this purpose, Quickbird image (15.09.2003) of a part of the Niepolomice Forest (south Poland) was used. Crown distribution and development of stands are related to the age and height of trees. The cross validation between different measurements clarify the potential of VHR data to achieve a reliable level of crown density estimation. This allows extrapolating the model over the whole forest area. For the forest management, stand age, species and height are the attributes incorporated in the SILP database (Polish State Forest). However, for the evaluation of successful growth per stand, crown density estimations are crucial and can change abruptly in a few growing seasons due to calamities. Only remote sensing methods and geoinformation techniques can assure low-cost and quick data collection over large forest areas and they are crucial when the needed information is reliable enough to provide decision support.