We produce high-resolution databases of all above-ground features – buildings, vegetation, and artificial engineering constructions.
Depending on the project goals and requirements, they can be represented as 2D shapes or 3D objects with the assigned attribute of absolute or relative height.
Buildings and bridge shapes and tree polygons are created using an automated production process of object recognition from high-resolution multispectral satellite images. Our Machine Learning methods developed based on Convolutional Neural Networks and the Deep Learning Techniques allow to extract 3D and 2D shapes for the entire country fast and at a high accuracy level.
- 95% of buildings with area > 25 sq.m matched automatically through the ML algorithm
- 100% coverage completeness is achieved due to manual post-processing and data validation
- X,Y, Z accuracy: ± 3m
The vegetation is diverse in nature and can be presented differently in the satellite image depending on various types, colors, heights, and seasons.
Our experts accomplished the training of neural network models using the training set comprised of 30 thousand objects of various vegetation patterns worldwide. This made it possible to achieve a high level of accuracy - 95-98% for test areas.
Obviously, such accuracy cannot be reached for the real data but due to the mandatory quality check process and manual data verification, the final models are very close to reality.
3D buildings and 3D Trees geodata along with related Digital Terrain Model (DTM), Clutter (LandUse) model, and basic vectors are perfectly suited for 5G networks deployment both in cities or high-urbanized areas and suburbs as well as for countrywide network design.
Accurate 2D and 3D buildings and vegetation are substantial components of multiple applications:
- WEB/GIS services
- urban and cadaster planning
- environmental and utility planning
We support all your projects and have products conforming to your requirements and budget.