Villages, lakes, coniferous forests: Trentino is easily recognizable from space.
You can see that from the images and maps of Trentino published on the website of the European Space Agency to present an innovative approach developed by Claudia Paris and Lorenzo Bruzzone of the University of Trento.
The two presented a new approach to integrate information provided by temporal series of satellite images and data automatically extracted from outdated land-cover maps, in the framework of a research project funded by the European Space Agency (Esa) and coordinated by the remote sensing lab (RSLab) of UniTrento.
Their efforts led to reliable maps created from a distance of 786 kilometres, which achieved an accuracy of over 93%.
The novel approach meets the demand for rapid, continuous and accurate land monitoring to examine the evolution of land cover on the Earth surface due to human activities and natural phenomena, and is part of a scientific programme launched by the European Space Agency.
The project's results were presented last week at Esa's Living Planet Symposium and were described in “A Novel Approach to the Unsupervised Update of Land-Cover Maps by Classification of Time Series of Multispectral Images”, an article by Claudia Paris and Lorenzo Bruzzone of the University of Trento and by Diego Fernandez-Prieto of the European Space Agency, published in “IEEE Transactions on Geoscience and Remote Sensing”. The European Space Agency appreciated it enough to re-launched it on its website illustrating progress made on the images and the relevant maps of Trentino.
This work is a further step forward in the use of satellite technology integrated with the most recent AI developments for the systemic and accurate update of land information.
The press release includes comments from Lorenzo Bruzzone – full professor at the Department of Information Engineering and Computer Science of the University of Trento, where he is also the director of the Remote Sensing Laboratory (RSLab) – and from Claudia Paris.