An automated and adaptable approach for characterizing and portioning cities into urban structure types

Abstract

Recently a growing number of investigations is dealing with the characterization and partitioning of urban agglomera-tions into urban structure types (USTs) based on remote sensing data. Since the USTs of interest are usually chosen with respect to the research question, application and type of urban agglomeration there is a need for a flexible and adapt-able approach for automatic UST classification. In this study we identify the commonalities of published approaches and derive requirements and tasks to deal with in UST classifica-tion. Based on this, we focus on the development of a UST classification system that is highly automated, flexible and adaptable to enable a wide applicability.