2005 年 49 巻 p. 1429-1434
In recent years, vegetation grows thickly in river channels even on originally gravel river beds. Riverine vegetation is one of the most important parameter for flood control, biodiversity conservation and so on. Consequently, it is quite important to develop efficient methods for monitoring these vegetation.
This research evaluated high resolution satellite remote sensing and classification trees methods for classification of riverine vegetation in the midstream of Tama river. The estimated vegetation classes were compared with the vegetation classes in the existing actual vegetation map and the accuracy were evaluated. These results suggest that the method having a combination of high resolution satellite IKONOS data and classification trees is useful for monitoring riverine vegetationin riparian ecosystem.