I would like to know the difference’s of the camera models, especially between fisheye and brown.
Comparing the resulting files there’s a lot of difference in the amount of data, the result from brown has a lot more data.
Are rectilinear wide angle lenses good for photogrammetry?
The models need to be appropriate for the lens used to capture the data. They’re a means to seed the lens coefficients and reconstruction.
I have some of them documented in our docs:
If you choose the wrong lens model for a certain set of data, you’ll get wildly varying reconstructions (if it doesn’t fail).
They should be!