Naslov
Detecting mistletoe infestation on Silver fir using hyperspectral images

Autori
Ančić, Mario ; Pernar, Renata ; Bajić, Milan ; Seletković, Ante ; Kolić, Jelena

Izvornik
IForest - biogeosciences and forestry (1971-7458) 7 (2014); 85-91

Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni

Ključne riječi
remote sensing; hyperspectral images; spectral signatures; mistletoe

Sažetak
Silver fir is one of the most important commercial and ecological species in Croatia, and is currently the most damaged tree species in Croatian forests. It is increasingly infested by white mistletoe, whose relevant consumption of water and nutrients contributes to physiological weakening of the tree. Early mapping of mistletoe infestation based on remote-sensed methods may overtake current limitations of standard field methods and their high costs in the assessment of tree health over large areas. The main goal of this study was to develop an efficient and reliable method for mistletoe detection based on hyperspectral images and supervised classification methods. Surveying was performed in a mountain area of Croatia characterized by beech-fir forests using a hyperspectral scanner. The Spectral Angle Mapper (SAM) classification with 5° threshold value proved to be the best classification method. The first “spectral signatures” - endmembers for fir and mistletoe, were determined and entered into a spectral library database. Using the above approach, a reliable picture of the spatial distribution of mistletoe-infected silver fir trees over the whole studied area was obtained. Results were validated by aerial surveying with a non-pilot aircraft from ~30 m above the crowns. The approach used here may allow a cost-effective mapping of the distribution and intensity of mistletoe infection in silver fir, and can be considered an helpful tool for forest management and planning, forest protection and harvesting.

Izvorni jezik
Engleski

Znanstvena područja
Geologija, Biologija, Šumarstvo

POVEZANOST RADA