Using Convolutional Networks and Satellite Imagery to Identify Patterns

Using Convolutional Networks and Satellite Imagery to Identify Patterns in Urban Environments at a Large Scale

Adrian Albert, Department of Civil and Environmental Engineering, Massachusetts Institute of Technology, MIT

Urban planning applications (energy audits, investment, etc.) require an understanding of built infrastructure and its environment, i.e., both low-level, physical features (amount of vegetation, building area and geometry etc.), as well as higher-level concepts such as land use classes (which encode expert understanding of socio-economic end uses). This kind of data is expensive and labor-intensive to obtain, which limits its availability (particularly in developing countries). We analyze patterns in land use in urban neighborhoods using large-scale satellite imagery data (which is available worldwide from third-party providers) and state-of-the-art computer vision techniques based on deep convolutional neural networks. For supervision, given the limited availability of standard benchmarks for remote-sensing data, we obtain ground truth land use class labels carefully sampled from open-source surveys, in particular the Urban Atlas land classification dataset of 20 land use classes across 300 European cities. We use this data to train and compare deep architectures which have recently shown good performance on standard computer vision tasks (image classification and segmentation), including on geospatial data. Furthermore, we show that the deep representations extracted from satellite imagery of urban environments can be used to compare neighborhoods across several cities. We make our dataset available for other machine learning researchers to use for remote-sensing applications.

More on

KDD2017 Conference is published on


  1. If we believe that biodiversity is valuable and we want to decrease the rate at which species are driven to extinction through loss of habitat, we could charge fees to landowners who use the landscape for purposes that interfere with or destroy wildlife habitat. Sharing fee proceeds to all people will ensure that the policy does not make those who are least well-off even worse off.

    Fees for monoculture would discourage waste of food and would promote methods of producing food that cause less impact on the environment. Fees for paving would discourage urban sprawl and would promote car-sharing and public transit.

    A random poll could reveal when most people think that rates of destruction of wildlife habitat and of encroachment on wilderness areas are not excessive. When conditions in the world match what most people want, the fees will be at the right amount. THEN we can say we live in a democracy.

    Biodiversity as a Public Good: