Navigating urban canyons: EU project combines satellite signals with camera images
Buildings and trees disrupt GPS and similar systems in cities. The EU project egeniouss, involving TU Braunschweig, adds images of the surroundings – with up to 50 times more accurate positioning in urban tests.
Archive image · Photo: Brunswyk / Wikimedia Commons, CC BY-SA 3.0 (cropped and resized)
In cities, buildings, trees and other obstacles block or reflect the signals of satellite navigation systems (GNSS), making positioning inaccurate. The European research project egeniouss, in which Technische Universität Braunschweig takes part, therefore combines satellite navigation with images of the surroundings and data from other sensors. This is meant to keep positioning working even when satellite signals alone are not enough. On 15 October, after three and a half years of work, the consortium presented its results.
A key result, according to TU Braunschweig: in urban tests, visual localisation achieved up to 50 times higher accuracy than conventional satellite-based methods. This was shown in three applications – bicycle navigation with the Naviki app, surveying with the QField app, and drones.
egeniouss was coordinated by the AIT Austrian Institute of Technology, with partners from Austria, Germany, Spain and Switzerland. The total budget was around 3.4 million euros, of which about 3 million comes from the EU's Horizon Europe programme. At TU Braunschweig, the Institute of Geodesy and Photogrammetry and the Institute of Flight Guidance took part.