I am a researcher studying PPK modules. For a balanced and ideal UAV flight, the roll and pitch values are expected to be less than 5 degrees. However, in practical scenarios, this is not always achievable.
If the M2 module is connected to a camera, how is the tilt compensated in such cases? When the camera captures a photo at an angle, I believe the M2 module provides the coordinates of the antenna’s phase center. I understand that photogrammetry attempts to correct camera poses, but if 50-60% of the flight involves significant tilt, can photogrammetry handle it effectively?
Additionally, how does the M2 module address such situations?
Good question! Reach M2 is responsible for providing an accurate position during the flight. Usually, this kind of compensation is on the drone’s side, and it can record the pitch, roll, and yaw values to the image’s metadata, which can then be used in the photogrammetric software.
Yes, indeed. But Reach M2 isn’t tilt-compensated overall. Alternatively, using it with an external IMU sensor may be possible. However, I can hardly share guidance for it as we haven’t tested such a setup ourselves.
Hi Amit,
If I understand properly, you have to differentiate two different issues here.
How to accurately measure (and use) the position of the drone
How to resolve a photogrammetry
For (1), as Kornel mentions, M2 does not provide a tilt compensated position, but the location of the phase center. If what you need is to follow the waypoints of the navigation, that won’t be a problem: An offset of a few centimeters will pass unnoticed.
For (2), the photogrammetry software uses all the data available to refine an accurate location of the camera location at the moment each image is taken. GPS location (any yaw, pitch, roll) of the drone are used as initial conditions for the computation. The contents of the images and the coherence between different images will provide enough data to refine the real camera parameters, that is, the real position/pitch/… but also the geometry of the camera itself. Initial conditions may be or may not be complemented with ground control points, points that are accuratelly meassured in the world and serve as inputs to the algorithm.
In summary, an accurate GPS location is desirable but not the only factor required. You can work with a GPS that provides locations with multiple meters of error, but if you have enough images with the right overlap and a set of ground control points, the solver will just work. If you use RTK/PPK, your result will gain in accuracy and the solver will have an easier task to complete. In general, photogrammetry software is surprisingly robust.
