Geoinformation, Environment and Infrastructure

We have been working on map creation, map analysis and geographic information systems for years in our Maps & Navigation field. The AI-based analysis of aerial images builds on this. We work for public administrations, infrastructure operators, and engineering and planning offices that want to use geodata in a structured and cost-effective way.

From the aerial image to the geodata base

Cities and regions have large image collections. Their value remains unused as long as every analysis is manual work. AI-based segmentation automatically generates object layers from an aerial image, for example for buildings, trees and vehicles. The layers can be shown and hidden individually, checked and processed further. In this way, an image becomes an analysable, reproducible geodata base.

Automatic detection is only the first step here. Specialists refine the result interactively, and every correction improves the model for the next run. We describe the methods under GeoAI & Spatial Intelligence.

Questions the method is suited for

The results serve as a data basis for planning, monitoring and reporting obligations. They can be integrated into existing geographic information systems, including in your own data centre.

Maps and digital twins

We provide individually prepared map data in two and three dimensions, in the browser or via a dedicated map server. The basis is freely available map data. It is supplemented with cadastral data, aerial and satellite images, and object data obtained from sensor data.

A digital twin becomes outdated when nobody maintains it: construction sites appear, vegetation grows. That is why we also use measurement drives as a source. What a vehicle detects along the way is located spatially and entered into the model. One example is the trial in a district of Chemnitz: AI-Based Environment Perception for Autonomous Public Transport.

How to get started

The easiest way is with a sample area: together, we analyse a section of your aerial image data and use it to show what can be detected and with what quality. We describe the entire process from the requirement to operation under Visual and Spatial Data Analysis, and the annotation tool under Acotate 2.0 AERO.