Terraloupe – Where AI meets Aerial Imagery

Aug 13, 2019

How does AI help to analyse aerial imagery and create accurate geo image data that can then be used to support the decision-making process in businesses? The Munich-based company TerraLoupe explains in a guest article how they do exactly this, opening up new prospects for numerous industrial applications.

TerraLoupe was founded in May 2015 and is a German startup for Artificial Intelligence (AI). At the centre of their operation is the belief that information is the new "gold" and understanding accurate geographic data can fundamentally improve decision-making. 

In the age of digitisation, vast amounts of data are being generated. Nevertheless, their analysis is often not mature. TerraLoupe's core competencies include automatic object recognition from aerial photography using deep learning techniques. Their deep learning algorithms allow users to automatically deliver accurate geospatial information to large geographic areas quickly and cost-effectively. With this, TerraLoupe can provide high-resolution, intelligent, and bespoke maps that pave the way for a digitised world. The technology thus allows the user the availability and optimal use of digital geodata.

With the help of 2D and 3D models, TerraLoupe create a "digital twin of the environment" in order to make other and, above all, new technologies possible. Be it autonomous mobility or smart cities – they are strengthening the applications of the future. Because a self-driving car must not only perceive immediate obstacles, but also recognize the traffic signs and know about road conditions know. 

The complete digitalisation of the environment through Deep Learning makes it possible to link a larger amount of information to a geocoordinate and thus make it quickly accessible in a medium. TerraLoupe's versatile object recognition supports companies in countless areas such as autonomous driving, the real estate industry, public transport and insurance. They count DAX corporations as well as international companies as their partners. 

Acquiring aerial photography has been one of the biggest challenges so far. Often these are licensed and can only be purchased at very high prices. Now this has been facilitated by partnerships and favorable reception options. Thus, future challenges arise more through the appropriate selection of cooperation partners to further develop the platform. 

Thinking of the future, the company envisions a digital environmental model that can be used for all applications, from air traffic for autonomous air taxis to virtual reality urban planning. They focus on the autonomous mobility market and will continue to develop their technology in this direction. TerraLoupe want to create a significant amount of data content for smart maps, so they are working step by step to deliver the big picture where autonomous systems don't remain on the ground but also take into consideration the height of buildings or trees for possible deliveries by drones or passenger transport. With their partners on board, TerraLoupe are a geodata provider that enables autonomous mobility. But of course, there is still much to do to digitise the whole world.