Impervious surfaces identified by artificial intelligence in Helsinki Metropolitan Area



Land cover mapping of impervious surfaces in Helsinki Metropolitan Area using AI. The pixel values represent the probability that the pixel is impervious, and they range between 0 and 254. Pixel values above 127 can be considered impervious. The spatial resolution is 20 cm and coordinate system is ETRS-TM35. Pixel classification is based on orthophotos and point clouds from National Land Survey of Finland (2020). The convolutional neural network used in production was developed by Scalgo ApS.

näytekuva

Short documentation (Mapping imperviousness for Greater Helsinki Region).

Download resources

Additional information

Published: 03.03.2022
Updated: 03.03.2022
Additional information:
Categories: Maps Environment and nature
Keywords: land cover impervious surface asphalt concrete pavement artificial intelligence AI neural network ortophotographs
HSY-botti