Imperviousness: recording and assessing sealed surfaces from high-resolution aerial photographs
Sealed areas such as paved traffic areas, car parks, industrial areas, airports and buildings lead to environmental problems such as increased surface runoff, reduced groundwater recharge and overheating. The localisation and quantification of these areas from aerial photographs is of great importance for the evaluation of land use and environmental impacts for planning offices, environmental authorities, municipalities and environmental associations. The company mundialis has developed a sealing processor that uses machine learning to analyse digital orthophotos (DOPs) and satellite images and recognise different surface classes.
The processor is based on remote sensing techniques and uses OpenStreetMap, ALKIS and Sentinel-2 data for automated classification. The recognised surface classes include sealed surfaces (roads and paths), buildings, unbound soil, water, forest vegetation and non-forest vegetation. The results can be provided in common geodata formats or as an OGC web service.
The sealing processor is currently available for the German federal states that provide the necessary geodata as Open Data. The extension to other federal states depends on the availability of input data.
As a future extension, mundialis is working on a “Street Surface Mapping” project that refines the existing sealing classification and identifies further surface categories within sealed surfaces (roads and paths), such as asphalted, paved and concrete roads, paved paths and car parks. This project is also based on deep learning methods and the analysis of aerial photographs and infrared data.
Have we caught your interest? Then contact us directly: email@example.com