作者: Demetris Stathakis , Konstantinos Perakis , Igor Savin
DOI: 10.1080/01431161.2012.687842
关键词: Segmentation 、 Satellite 、 Cartography 、 Context (language use) 、 Data type 、 Remote sensing 、 Index (economics) 、 Normalized Difference Vegetation Index 、 Vegetation 、 Computer science 、 Remote sensing (archaeology)
摘要: Urban populations are expanding rapidly and so cities. Remote sensing offers a convenient means of monitoring this expansion as it covers period 40 years in the case LANDSAT satellite. In some parts globe, is probably only viable due to lack other types data. order monitor expansion, first, urban land has be separated from land-cover types. Although can done by standard classification processes, much more efficient establish an index UI analogous widely used normalized difference vegetation NDVI for vegetation. Existing efforts such reviewed compared common context. Following this, novel, introduced. The calculation new straightforward, based on combining with built-up index. results promising efficiently segment areas, even presence excessive bare land. proposed method evaluated two test sites selected different scenes. valid sensors same bands those LANDSAT.