Application of High Spatial Resolution Satellite Imagery for Urban Environment Mapping

The aim of the study is to develop automatic procedures for the identification of urbanized and not urbanized areas using high spatial resolution remote sensing data. We examine the potential of panchromatic and multispectral Quickbird images for the classification of different urban areas in the Province of Milan (Italy). Satellite data were initially pre-processed by using an orthorectification procedure with ground control points in order to achieve a high geometrical accuracy. Geometrical accuracy was evaluated by means of the analysis of the root mean square positioning error that was found to be from 2.5 to 5 m in the different images. An object-oriented classification based on a multi-resolution segmentation technique was applied by means of the e.Cognition software, in order to obtain the classification of urban land cover types. This methodology allowed the creation of thematic maps with several intra-urban land use classes, which can be used as a cartographic background in urban studies. The results confirm that the classification of Quickbird images is a powerful tool for urban development, environmental protection, policy making, risk assessment and planning.