The Postcode Context Classification provides a national measure of urban spatial structure derived from high-resolution satellite imagery and cutting-edge convolutional neural network (CNN) techniques. By harnessing the power of the European Space Agency’s Copernicus Sentinel 2 satellites, combined with georeferenced postcodes, this data product offers unparalleled insights into the built environment across Great Britain.

The dataset is an invaluable tool for urban planners, researchers, and policymakers, providing detailed classification of diverse urban and rural contexts. It presents a unique approach to urban spatial analysis, leveraging advanced machine learning and satellite data, marking a significant advancement in the field of urban studies, and opening new avenues to understand and shape the evolving urban landscape.

The dataset presents a new method of measuring local context that could be flexibly applied within different settings to meet several definitions of neighbourhood. While the method is implemented within the context of Great Britain, given the global coverage of satellite imagery, the approach can also be applied to any location in the world. The limits of the technique are centred around the resolution of the satellite data used and the interaction between the geography of the input data and the learned structure.

For details about the methods that were used to create the Postcode Context Classification, see the open access paper linked below.