ABSTRACT
References .............................................................................................................. 381
Urban areas are regarded as a very complex landscape in terms of the diversity of land cover types and the shape and pattern of various urban features. Accurate and up-to-date information on urban land cover is a challenging task but of crucial importance for urban planning, environment protection, and policy-making. Urban extent extraction and land cover mapping have been studied using a range of remotely sensed data and algorithms. Gamba and Herold (2009) summarized the topic with a special focus on global monitoring. High-resolution optical imagery and object-based approaches are often used for urban applications at the local level, while regional or global analysis generally exploits moderate-or coarse-resolution optical images and pixel-based methods (Gamba and Lisini, 2013). In addition to optical data, synthetic aperture radar (SAR) systems have been playing an increasingly important role in urban analysis due to their ability to acquire images day and night in all weathers and to the fact that the number of advanced SAR systems in operation is increasing (Rogan and Chen, 2004). However, the specific imaging characteristics of SAR systems and the existence of speckle noise make the interpretation of SAR images in urban areas generally more difficult compared to the analysis of optical images. Nevertheless, SAR data have been investigated for urban extent extraction and land cover mapping with promising results (Ban et al., 2010, Gamba et al., 2011; Hu and Ban, 2012, Niu and Ban, 2013).
