Hullo All,
I'm having some trouble figuring out a workflow to stitch photos from
different lenses. This is something that I wanted to learn about, so
as usual I decided to try to prepare a tutorial on the subject,
figuring that the best way to understand it is to be able explain it
to others.

I am using Hugin 0.8 with Autopano-SIFT-c.
I have no trouble at all when using jpeg input images. Using the
Assistant, Hugin gets the EXIF data and alignment and stitching goes
well, automagically, and I get a good result.

The troubles start when using input images without EXIF data.

I tried using tiff copies of some jpeg images that had worked well.
The steps I have used are as follows;
1. load first tiff image, enter a guess at the HFOV (used 60 deg,
actual value from EXIF is 63.4 deg)
2. load 2nd tiff, go to 'Camera and Lens' tab, select 2nd image and
select 'New lens', then select 'Load EXIF' and enter 40 deg (actual
value is 42.6 deg)
3. 'Optimizer' tab set to optimize 'everything', 'Exposure' tab set to
'Low dynamic range, variable white balance', 'Stitcher' tab, select
'normal output, JPEG'.
4. Select '2.Align', reports good fit, mean error=2.3, max=5.4 (using
jpeg images mean=1.5, max=4.7)
5. 'Camera and lens' tab shows optimisation didn't correct/alter the
guessed HFOVs
6. Reset 'Field of View' on each image, rerun optimizer. HFOVs reset
to 59.9, 40.1, mean error=1.4, max=2.9, result shows image 0 looking
good but image 1 is a bit distorted, not too bad but certainly not as
good as the result using the automated process with jpeg images.

I note that providing an HFOV to Hugin results in Hugin setting the
focal length with a focal length magnifier of 1.0. During one of my
experiments I provided the focal length magnifier (6.08) along with an
estimate of the HFOV and the resulting stitch was better than when
just providing the HFOV.

The above is a summary of the steps that provide a result, but I have
tried numerous other approaches to this and it is beginning to look to
me like there isn't a sound repeatable process that one could use for
images that don't have EXIF data, unless one can make really good
guesses at the HFOV, Focal length and focal length magnifier.

This leads to some questions.

If one has images that don't have EXIF data, how good do the initial
estimates of HFOV need to be to ensure a good result?

Is there a set of foolproof steps one can use when input images do not
have EXIF data?

Cheers,
Terry


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"hugin and other free panoramic software" group.
A list of frequently asked questions is available at: 
http://wiki.panotools.org/Hugin_FAQ
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at http://groups.google.com/group/hugin-ptx
-~----------~----~----~----~------~----~------~--~---

Reply via email to