Determining the Relationship Between Census Data and Spatial Features Derived From High-Resolution Imagery in Accra, Ghana

Abstract

Remotely sensed-based estimates of dwelling and population characteristics can provide timely and spatially explicit information for urban planning and development in emerging cities. This exploratory analysis quantifies spatial features of built-up areas derived from high-resolution satellite imagery and directly relates them to census-derived variables in Accra, Ghana. Spatial features are image metrics that analyze pixel groups in order to describe the geometry, orientation, and patterns of objects in an image. By using spatial features in an urban setting, city infrastructure variations, such as roads and buildings, can be quantified and related to census variables, such as living standards and housing conditions. To test the associations between spatial patterns and demographic variables, five spatial features (line support regions, PanTex, histograms of oriented gradients, local binary patterns, and Fourier transform) were quantified and extracted from the imagery, and then correlated to the census variables. Findings indicate both spatial features and spectral information (such as NDVI) correlate strongly with standards of living such as population and housing density. Results from this study suggest that spatial features derived from satellite imagery can be used to help map socioeconomic characteristics within the city of Accra, Ghana, and that this methodology may be transferable to other developing cities.