期刊名称：GI_FORUM - Journal for Geographic Information Science
电子版ISSN：2308-1708
出版年度：2015
页码：406-415
DOI：10.1553/giscience2015s406
出版社：ÖAW Verlag, Wien
摘要：Automated and visual approaches for the monitoring of refugee or IDP camps based onsatellite data are very important as independent information sources, especially for insecureand remote areas. Nevertheless, monitoring based on satellite data always has a certaindegree of uncertainty, e.g. due to data quality, complexity of the area of investigation, seasonalpheonological problems, or algorithmic limitations. Within this paper, we aim toquantify one of these limiting aspects: the factor of vegetation (i.e. tree) growth and itseffect on multi-temporal dwelling monitoring, hampering the identification of dwellings onthe ground. For the refugee camp Djabal, Chad, we found that 2506 dwellings (25 %) of2010 are at least partly affected by tree growth three years later (2013), which is influencingautomated extraction methods, as well as visual interpretations. 395 of these dwellingswere completely covered by vegetation and vegetation shadow, and were therefore notdetectable anymore. Taking this factor into account, the decrease of dwellings between2010 and 2013 is potentially lowered from 10 % to 5 %.
