Application of phenology to assist in hyperspectral species classification of a northern hardwood forest
Note: imported from RIT’s Digital Media Library running on DSpace to RIT Scholar Works in December 2013. Physical copy available through RIT's The Wallace Library at: QH544 .S77 2005
Tree species have unique spectral reflectance patterns that allows them to be both compared to other objects and to other types of trees. Increasing the spectral separation of such images may assist with surveying and forestry inventories. In past studies, most classifications were done with summer leaves, which darken and become very similar shades of green. This study utilized the phenology of trees to investigate how the changing colors of young or senescing leaves may assist in species classification based on aerial images. Images were taken of the Hubbard Brook Experimental Forest, which is mainly dominated by sugar maple (Acer saccharum Marsh), beech (Fagus grandifolia Ehrh.), and yellow birch (Betula allegheniensism Britt). Classification of stands of same-species trees was attempted using spring hyperspectral images containing bands from fall RGB color photos. The combination of high-resolution RGB photos and lower-resolution hyperspectral data was found not to increase the spectral separation when combined.