Slum mapping in polarimetric SAR data using spatial features

Spatial patterns of deprivation in cities of the global south in very high resolution imagery

“…Thus image texture features showed a much better transferability within and across cities for mapping deprivation, which was also confirmed by other studies (e.g., Wurm et al, 2017). To combine texture and spectral image feature, machine learning algorithms are of high utility, being able to deal with a large number of input features for producing more accurate results than standard parametric classifier.…”

“…Fourth, to search for methods that allow to capture the diversity of deprived areas, moving away from the 'one size fits all' approach. The first principle is used within OBIA studies (e.g., but most machine leaning studies use either pixel-level outputs, regular tiles or windows (e.g., Wurm, Taubenböck, Weigand, & Schmitt, 2017). The second and fourth principles have been largely ignored by most studies on mapping deprived areas.…”

“…Neighbourhood boundaries might be defined by using additional data layers such as OSM data (e.g., blocks) or image based features (e.g., texture), for instance, to better capture low density built-up areas. In addition, more attention should go to the inclusion of small pockets of deprivation, which are poorly covered in most mapping approaches (e.g., Wurm et al, 2017). Leaving out the small pockets bears the risk that we ignore one of the most deprived groups in urban areas, which commonly has no voice in relation to urban planning and management and possible upgrading schemes.…”

“…The utility of VHR imagery covering large areas is particularly relevant for complex megacities with rapid changes. With respect to the megacity of Mumbai, several studies have shown the potential of VHR imagery to map urban land uses (e.g., Kamini, Jayanthi, & Raghavswamy, 2006), and particularly deprived areas (e.g., Taubenböck & Kraff, 2014;Taubenböck & Wurm, 2015;Wurm et al, 2017) that house a large share of the population. Nevertheless, to date, little research has quantified their specific morphological characteristics.…”

A Critical Review of High and Very High-Resolution Remote Sensing Approaches for Detecting and Mapping Slums: Trends, Challenges and Emerging Opportunities

“…However, this study utilized only few classes for classifying the land use land cover, thus requiring additional research for evaluating the suitability of this approach in more complex urban environments. It is also worth noting that this study is one of only a few who exploited radar imagery for mapping slums (e.g., [98][99][100]). …”