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：2000
卷号：XXXIII Part B7(/1-4)
页码：1601-1608
出版社：Copernicus Publications
摘要：Assessing land cover changes in tropical rain forest areas is an issue of worldwide importance. Multi-spectral, higher spatial resolution satellite data are often the only data source for generating the necessary change information. Applying conventional remote sensing processing methods, however, induce uncertainties for both the spatial and thematic components of land cover changes in tropical rain forest areas. The induced uncertainties are not a problem of mixed- pixels, but concern: (a) the relation between the aggregation level of change processes in tropical rain forests and the measurement resolution of the remote sensing data used to observe these change processes, and (b) the problem of handling heterogeneous object classes with fuzzy extents at supra-pixel level. To model the object-related induced uncertainties the concept of a process-driven semantic approach is described in this article. It is based on the use of aggregate sets occurring at supra-pixel level to model changes of heterogeneous spatial objects with fuzzy extents. A conceptual framework on aggregate sets, aggregate classification and aggregate change assessment is defined and related classification and change assessment models are described. The methodical development to process remote sensing imagery at supra-pixel level for different aggregation levels of land cover changes, is a further step to model real-world phenomena
