Job Description:
Vision-based localization for GNSS-denied navigation
In both manned and unmanned aerial vehicles, GPS is a critical component for navigation and positioning. However, GPS signals can be denied or degraded due to jamming, spoofing or other causes. This necessitates alternative self-localization methods to enable assured positioning, navigation, and timing (APNT).
One potential solution to this problem is vision-based navigation using map matching. This approach involves maintaining a database of georeferenced nadir images captured by aircraft or satellites. By comparing the features and landmarks in these images with the current view of the area over which the vehicle is flying, the system can determine its own position relative to local maps.
There has already been research on this topic showing that this navigation technique can provide performance comparable to GPS in detailed environments.
The aim of this thesis is to provide a methodology for the features matching and self-localization, to evaluate the performance and highlight potential drawbacks or criticalities of the proposed navigation technique.
Finally, a complementary study following the localization algorithm would consist in proposing a way to integrate this system into the Ego Vehicle navigation architecture and setting the foundations for sensor fusion with the Inertial Navigation System.
Seniority:
Primary Location:
IT - Varese - Cascina Costa
Additional Locations:
Contract Type:
Hybrid Working: