Hallöchen!
If I understand the Hugin projection model correctly, the distance
of a pixel from the image centre r (in real-world coordinates),
depending of the angle of the incoming light ray with respect to the
optical axis, is
r(alpha) = f * (a*r^3 + b*r^2 +c*r + 1 - a - b - c) * tan(alpha)
where "tan(alpha)" is for rectilinear lenses, and being replaced by
other terms of other types of lenses.
However then, the focal length f is wrongly defined. In the
paraxial approximation, where r and alpha --> 0, you get
r(alpha) = f * (1 - a - b - c) * alpha
but it should be
r(alpha) = f * alpha.
So, Hugin moves some of the distortion into the focal length,
getting the focal length wrong by a factor of (1 - a - b - c). I
suspect that the FOV is equally affected. That said, the projection
formula as a whole does a correct mapping.
Is this historical burden from the PT project that we have to live
with, or is this focal length thing "far enough" form the core that
it is worth a bug report? I mean, old FOV values would become
incompatible if this was fixed. On the other hand, one could
maybe only fix the displayed "focal length" in the GUI, as I think
this is not written to the files anyway.
I ask this because I care for the Lensfun project at the moment, and
it inherits all the formulae from Hugin. With all their quirks.
;-)
Tschö,
Torsten.
--
Torsten Bronger Jabber ID: torsten.bron...@jabber.rwth-aachen.de
or http://bronger-jmp.appspot.com
--
A list of frequently asked questions is available at:
http://wiki.panotools.org/Hugin_FAQ
---
You received this message because you are subscribed to the Google Groups
"hugin and other free panoramic software" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to hugin-ptx+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/hugin-ptx/87ppl979b9.fsf%40physik.rwth-aachen.de.
For more options, visit https://groups.google.com/d/optout.
