Autonomous drone navigation with 3D processing

Spacemetric is investigating image-based methods to derive 3D information for drone navigation at low altitudes independent of GPS over longer distances in forested landscapes. The project is taking place in the framework of the innovation programme of the Swedish Defence Materiel Administration (FMV).

An onboard camera can provide measurements of drone movements that can be combined with other information, such as inertial measurements, to determine its position. Over longer distances, corrections for positioning drift can be sourced from additional information, such as aerial images or satellite images. Remaining challenges can be mitigated through the application of 3D information.

Spacemetric has previously produced a 3D-based positioning method which was evaluated using aerial images in experiments over Swedish forests. The new project will evaluate real-time generation and use of such 3D data from drone images or video using software from I-CONIC.

Previous
Previous

Imagery and AI analytics for Carmenta

Next
Next

Dating forest fellings with radar imagery and AI