Our Building Detection service allows to detect the area occupied by buildings and artificial structures inside the geographical area under study (area of interest, AOI) based on their canopy from high spatial resolution satellite imagery, using advanced, state-of-the-art proprietary AI models, developed by PERIOPSIS.
Access analytics powered by emerging technologies
Satellite Imagery
High spatial resolution satellite imagery, @0.3-0.5 meters/pixel, capable of identifying building areas regardless their size or shape.
Artificial Intelligence
Proprietary AI models trained on extensive datasets to detect building areas with high precision (94%), high recall (99%) and high Intersection over Union (0.89).
Validation geographic area: Island of Cyprus (9,000 square kilometers)
Output: Map of all building areas as a GIS layer
FAQs
Our Building Detection service can detect trees with high precision (94%), high recall (99%) and high Intersection over Union (0.89). You can check our blog where we describe our service and its overall performance. We consider our results as state-of-the-art, considering both research and commercial competition.
The frequency and timeframe of surveys depend on your needs and resources. We recommend annual surveys, ensuring efficient urban monitoring and taxation updates. One-off surveys can also cover planning and baselines for shorter projects. Surveys combining recent and historical data are recommended to identify trends in development activity.
We can provide estimated addresses for each detected building using trusted third-party maps. Address matching quality depends on local data availability and coverage, so results may vary by region.
Simply sign up on the PERIOPSIS Portal and submit a new project following the guided procedure. There, you can select an area of interest (AOI) manually by drawing a polygon or by selecting a predefined polygon, in case you represent a regional authority. We handle your personal information with the highest standards of security and privacy, in full compliance with the GDPR.
Results are delivered as interactive maps and analytics via our portal. Other formats such as georeferenced vector files (e.g., GeoJSON, shapefile and KML) can also be arranged.
Our current output provides 2D footprints of buildings with area estimates. Height estimation can be considered as an additional element of a project—you can find more information in our relevant blog.
Yes, it is possible. Please contact us first at [email protected] so we can review your data. If you have imagery meeting our resolution and metadata requirements, we can run our models on it.
We currently fully support Cyprus and Greece. If your region is not supported but you still want to use our services, send us an email at [email protected].
This ranges between 15 and 30 days, depending on area size and weather (i.e., cloud) conditions.