A real-time visual localisation framework for satellite-based navigation


The EU project egeniouss makes accurate and reliable navigation available for everyone. Satellite-based navigation and positioning is the backbone for a multitude of services and even entire industries. But in cities, urban canyons interrupt and reflect satellite signals making navigation unrealiable und imprecise.

Additonally, satellite signals can be easily manipulated or blocked. In an age where positioning has become a truly democratised technology and autonomous systems are in the wings, the overreliance on a single source of geolocalisation poses certain safety risks but also hinders many professional and consumer applications to unfold their full potential. egeniouss will change that by bringing a visual localisation cloud service to the market.



With surveying-grade geodata in the background, egeniouss aims for an absolute accuracy of up to 10 cm* offering unparalleled performance for all sorts of location-dependent applications.

*depending on the applicable service level


Satellite navigation is prone to intended and unintended signal manipulation making autonomous systems difficult if not impossible to operate. egeniouss offers a redundant positioning solution enabling higher reliablity even in difficult receiver environments, such as urban areas.

Platform independence

egeniouss is a platform-independent cloud-based solution. It can run on smartphones, drones and will later be made available for other platforms as well.


Machine Learning

Machine Learning is an essential tool for optimising many components in egeniouss.

For instance, it does not only provide the semantic understanding of images but helps to make image descriptors more salient and supports estimating transformations between images in a sequence.

A variety of neural networks are deployed and continuously optimised to improve egeniouss’ performance over time.


egeniouss utilises highly accurate aerial and terrestrial mobile mapping geo data as reference for visual localisation. This data captures the world from different perspectives and hence enables visual localisation for a wide variety of platforms and operating modes.


egeniouss is designed as a cloud solution to facilitate scalability, reliability and performance as well as the management of a large amount of reference data. It also allows for offline operation through caching if no internet connection is available.


egeniouss is made available through a platform-independent API. This enables its convenient deployment to mobile phones and drones but also other devices and platforms in the future.


egeniouss is developed as an application-independent service highly advantageous for a wide array of location-dependent or location-reliant applications.

During the project runtime, egeniouss will demonstrate its performance and versatility with three complementary use cases in terms of reliability, accuracy and latency criteria – smartphone-based surveying with the industry-leading app QField, drone delivery beyond visual line of sight and in urban areas as well as cycle navigation in a dense city environment.

Crowd sourcing

As a visual localisation service, egeniouss needs to be up to date to function properly. While surveying-grade geo data is employed to update the reference data in the cloud in regular intervals, crowdsourcing is used as an opt-in feature to maintain the data integrity in shorter intervals.


The European GNSS Galileo and its augmentations, such as the High Accuracy Service (HAS), are the backbone of egeniouss. egeniouss’ multi-sensor navigation integrates code and carrier phase measurements as well as combinations thereof including HAS into a unified, general and generic, sensor-agnostic real-time trajectory determination module.

Geodesy, Photogrammetry & Computer Vision

Visual localisation is a challenging technology relying on the successful interplay of multiple scientific fields. In egeniouss, scientists from Computer Vision, Photogrammetry and Geodesy work closely together to build the algorithmic foundation of the technology.

Computer Vision algorithms and neural networks enable the understanding of the contents of images, techniques from Photogrammetry accurately associate images geometrically while methods from Geodesy are employed to integrate all measurements into a powerful estimation framework.

Evolution of European Policy Framework


European Green Deal


European Green Deal


European Green Deal


European Green Deal




Social Media

Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or EUSPA. Neither the European Union nor the granting authority can be held responsible for them.”