Extraction of buildings from quickbird imagery: What is the relevance of urban context and heterogeneity?

The high frequency and scope of spatial changes in cities demands ways of expediting the production and updating of large-scale geographic information. For that purpose, current and future very high spatial resolution satellite imagery (VHR) and semiautomated object-based image analysis methods may be an advantageous alternative to classical data sources and methods, i.e., aerial photography and photogrammetry. At the same time, the urban environment is becoming more complex and heterogeneous, possibly turning the feature extraction process more challenging. While much research has focused on developing, adapting and applying these approaches, less attention has been devoted to the interplay of data source (imagery), feature extraction methods, and geographic characteristics of the area under analysis. Lisbon, Portugal, is both a historical and modern city having a dynamic landscape, where increasingly diverse urban forms and materials coexist. This complex reality is possibly causing the feature extraction process from imagery to become more challenging. This study tests the semi-automated extraction of buildings from a QuickBird image in several urban study areas in Lisbon having different characteristics, and explores the impact of the heterogeneity of these features in the extraction process. Spatial metrics and spectral response are used to characterize types of buildings present in the study areas. Results show that the study areas display different levels of heterogeneity even for the same type of building and suggest that the quality of the extraction is affected by more factors than the complex variations in color/tone, composition and spatial configuration of target features.