XYZ optimization parameters

215 views
Skip to first unread message

Nicolas Pelletier

unread,
Apr 9, 2010, 7:20:57 AM4/9/10
to hugi...@googlegroups.com
Hi,

I'm playing around with one of the intermediate builds (5063) provided by this wesite (http://lemur.dreamhosters.com/hugin/).

In the image positioning (and optimization parameters) there are 3 new ones, namely X Y and Z.

Are these the "viewpoint" or tilt options (not sure what the name is)?

If it is, what is the status on those?

I've been playing with them trying to stitch a nadir on an old pano (when I didn't shoot them properly) with the hope that it could save it.

Currently, it has special effects on one of the images (with the larger values) which, when I look at the preview seems to "cut" par of the image (but still is properly positioned). This leaves hole in the output.

So, I only want to confirm if this is a CODE18, and if so, if there are any pointers to help correct this.

Thanks,

nick

Bruno Postle

unread,
Apr 9, 2010, 8:10:24 AM4/9/10
to hugi...@googlegroups.com
On 9 April 2010 11:20, Nicolas Pelletier <nicolas....@gmail.com> wrote:
> In the image positioning (and optimization parameters) there are 3 new ones,
> namely X Y and Z.
> Are these the "viewpoint" or tilt options (not sure what the name is)?
> If it is, what is the status on those?
> I've been playing with them trying to stitch a nadir on an old pano (when I
> didn't shoot them properly) with the hope that it could save it.

It is 'viewpoint correction' or 'mosaic', and it works very well but
in a restricted way. It assumes you have a series of photos of a
planar object like a painting or a mural taken from different
locations and directions. It then assembles them into a 'plane' that
is perpendicular to the view direction of the panorama.

So if you try it with a series of pictures of the ground, the 'nadir'
always ends-up in the middle of the view - This will be a little odd
with a spherical panorama. I'd recommend that at the moment you use it
to create a clean rectilinear 'nadir' image which you can then insert
into a 'normal' panorama project.

--
Bruno

Nicolas Pelletier

unread,
Apr 9, 2010, 8:52:50 AM4/9/10
to hugi...@googlegroups.com
Thanks for the info. I'll try it as a two step for the moment.

Will this work with a spherical at one point, or the algorithms don't permit that?

Thanks,

nick


--
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 hugi...@googlegroups.com
To unsubscribe from this group, send email to hugin-ptx+...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/hugin-ptx

To unsubscribe, reply using "remove me" as the subject.

Stefan Peter

unread,
Apr 9, 2010, 8:53:55 AM4/9/10
to hugi...@googlegroups.com
Hi

Bruno Postle schrieb:


> On 9 April 2010 11:20, Nicolas Pelletier <nicolas....@gmail.com> wrote:
>> In the image positioning (and optimization parameters) there are 3 new ones,
>> namely X Y and Z.
>> Are these the "viewpoint" or tilt options (not sure what the name is)?

...


> I'd recommend that at the moment you use it
> to create a clean rectilinear 'nadir' image which you can then insert
> into a 'normal' panorama project.
>

Could this statement be augmented with a small example or even a
tutorial? Or does this exist already but has managed to avoid my prying
eyes?

Regards

Stefan Peter

Bruno Postle

unread,
Apr 9, 2010, 7:45:40 PM4/9/10
to Hugin ptx
On Fri 09-Apr-2010 at 14:53 +0200, Stefan Peter wrote:
>Bruno Postle schrieb:

>> I'd recommend that at the moment you use it
>> to create a clean rectilinear 'nadir' image which you can then insert
>> into a 'normal' panorama project.

>Could this statement be augmented with a small example or even a
>tutorial? Or does this exist already but has managed to avoid my prying
>eyes?

Sorry, so far I've only used mosaic mode for the 'normal' case of
photographing a painting and haven't tried anything else.

--
Bruno

Bruno Postle

unread,
Apr 9, 2010, 7:45:32 PM4/9/10
to Hugin ptx
On Fri 09-Apr-2010 at 08:52 -0400, Nicolas Pelletier wrote:
> Thanks for the info. I'll try it as a two step for the moment.
>
> Will this work with a spherical at one point, or the algorithms don't permit
> that?

I don't know, libpano13 has more than one system for doing this, but
the one that is implemented in Hugin is the only one I have tried.

Nicolas Pelletier

unread,
Apr 9, 2010, 10:08:19 PM4/9/10
to hugi...@googlegroups.com
Is there a reference documentation, somewhere I could confirm if there are some tilt parameters, and what are their names?

I currently use command line scripts (python) to do most of the work, so if some parameters are available there, I'd try that also!

Thanks,

nick

Julien Schroder

unread,
Feb 1, 2013, 12:57:13 PM2/1/13
to hugi...@googlegroups.com
Is there a way to throw actual geographic coordinates in those parameters in a remote sensing context in order to improve the mosaic?

Bruno Postle

unread,
Feb 1, 2013, 2:33:29 PM2/1/13
to Hugin ptx
On Fri 01-Feb-2013 at 09:57 -0800, Julien Schroder wrote:
>Is there a way to throw actual geographic coordinates in those parameters
>in a remote sensing context in order to improve the mosaic?

In principle yes, if you think of each photo as a slide projector
projecting onto a flat plane, the XYZ coordinates are the position
of camera for each photo.

--
Bruno
Reply all
Reply to author
Forward
0 new messages