期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2016
卷号：XL-3/W4
出版社：Copernicus Publications
摘要：Information about alterations of vegetation compositions and structures within one growth period is required for different applications in ecology, nature conservation, and agriculture. Since plants change their spectral behavior during the growth period, a remote sensing-based classification is complicated, involving exhaustive field-work. Because a variety of plant communities can have a similar spectral signal for a single acquired image, a multitemporal classification is often utilized for class separations. However, this approach multiplies the required field work. But even if a multitemporal classification was successfully applied, the approach will be only limited transferable due to shifting acquisition dates and vegetation compositions. For monotemporal detecting of different materials, a spectral library is frequently utilized. In order to distinguish a variety of habitats at all stages of a growth period an available spectral library is not sufficient, due to the changing reflectance. Therefore, the concept of a phenologic library is introduced. A phenological library could be used to optimize the image acquisition of earth observation data in terms of detecting the dates of highest separability of classes. With repeated intra-annual spectral measurements of different vegetation types, this method can combine the transferability of spectral library approaches with the higher class differentiation of a multitemporal classification. Additionally, a phenological library is assumed to be transferable to other acquisition dates of other years
关键词：phenologic library; RapidEye; Very High Temporal Resolution (VHTR)
