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：2006
卷号：XXXVI-4/C42
出版社：Copernicus Publications
摘要：For flood risk modelling at watershed level, it is important to identify urban areas due to their soil waterproofing function. Very high spatial resolution images are now able to discriminate the elementary features which characterize urban landscape. Nevertheless, the traditional per-pixel classifier approach appears inadequate for mapping these landscape objects. We present an image analysis flowchart, based on the use of object-oriented software (eCognition). The maps obtained identify the urban landscape at different scale levels (urban areas and buildings – settlements) and at different organization levels (dense, discontinuous or low-density urban landscape). A particular effort has been made to assess the quality of the results (confusion matrix, Kappa coefficient), in relation to field data
