There have recently been remarkable advancements in EO state-of-the-art information extraction techniques using ML/DL approaches, including in the urban domain. However, local in-situ information for training machine-learning algorithms, Quality Assurance/Quality Control and validation are often limiting factors, due to their sub-optimal availability and coverage.
In general, there are a lack of appropriate in-situ data collections for most urban monitoring themes, since this is demanding both organisationally and financially. This is true particularly for remote or unsafe regions around the world and/or complex thematic features. Therefore, there is still considerable room for improvement in the organisation of such collections and thus a huge potential to further operationalise many urban mapping and monitoring services.
In this context, a novel scalable framework for EO-based information extraction, supported by crowdsourcing and gaming approaches, has been developed within the framework of the ESA supported GAME.EO project.
The approach is demonstrated using selected service cases for EO-based monitoring of slum areas (UN Sustainable Development Goal 11) in the development-support context. The tool is aimed at the actionable use of crowdsourcing and gaming tools to enhance current machine-learning algorithms for the identification, delineation and further characterisation of slum areas. The developed framework and tools are tested in cooperation with World Bank Group users and stakeholders from Water and Urban Global Practices.
Tools supporting crowdsourcing campaigns via gaming engagements can be used to mobilise and train volunteers to provide training data and to help extract required information in a more timely and accurate manner and with lower operational costs than would be incurred using standard data collection services. It demonstrates a real potential and an added value in terms of the synergies of crowdsourcing and EO-based information to support the development actors and cities’ research and operational activities globally.
