[hugin-ptx] Re: Towards a non-patented control point detector

[Pablo d'Angelo]

Bruno Postle wrote:

On Thu 11-Feb-2010 at 09:18 +0100, Harry van der Wolf wrote:


I'm not totally happy with the name panomatic is that is already an 
existing
name for an existing CP matcher/detector. Can't we call the panomatic 
binary

something like "cp_matcher" or "hugin_matcher" or "hugin_cp_matcher" or
anything else logical, not interfering and confusing with an existing
program.


'panomatic' is ok so long as it replaces the existing panomatic, and the 
panomatic webpage indicates this.


Currently, it should be a drop in replacement. A new name would be nice.

There is no way to build my extended package without the SURF 
descriptor, so its not yet suitable for general redistribution, some 
more small modification to the build scripts are necessary.


...but I think that ultimately this needs to be merged into Hugin and 
built as a Hugin tool, for two reasons:


Indeed, this is why I made the detector and descriptor as a separate 
library, which could be easily linked to any program.


ciao
  Pablo

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

[hugin-ptx] Re: Hugin's vignetting formula

[Pablo d'Angelo]

Robert wrote:

Hi,

I find this a bit dubious, since the input is a camera JPEG, and
therefore I would expect the manufacturer to compensate for sensor
gain and normalise it for display purposes (i.e., apply inv. gamma).


Cameras are made to make nice to look at pictures, not as scientific 
measurement devices. The manifacturers do various tricks to make nice 
looking photos. I'm also not sure what all the new "adaptive" functions 
like d-lighting etc. do. They might even apply a local changes/contrast 
enhancement etc. that is impossible to capture with a global transformation.



Okay, so all I have to do now is make sense of that EMoR paper by
those UOC guys. Hm, Hugin apparently uses only 5 parameters, but at
first glance in the paper and in 'emor.txt' I saw values for something
calles E(?), and curves f0-f25 (so quite a few more than just 5). The
PanoTools wiki does not say _anything_ about the parameters (the UI
does not even name them...), apart from "you are not supposed to guess
them". I do not want to guess either, but in order to become
compatible with Hugin, I have to find out their definition...


Just read the paper by Grossberg et. al. It explains exactly what that 
EMoR is. In a nutshell, its the PCA decomposition of a large set of 
camera responses. The first few dimension are enough to capture the 
typical variation of all cameras.


ciao
  Pablo

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

[hugin-ptx] Re: Sourceforge export controls

[Yuv]

On Feb 9, 5:23 pm, Bruno Postle  wrote:
> On Tue 09-Feb-2010 at 17:57 +, James Legg wrote:
>
> >Sourceforge is blocking users from some countries from downloading
> >files[0]. Recently they added the an option for project admins to remove
> >the block on a project-by-project basis if it doesn't do encryption[1].
>
> >I don't think Hugin, Panorama tools, Enblend, Luminance HDR, etc do
> >encryption, so could the project admins remove the blocks?
>
> Ok, done for Hugin and Panotools.

Somebody did this for Enblend as well (I logged in and it was already
done).
Yuv

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

[hugin-ptx] Re: Can I join?

[Yuv]

Bonjour Antoine

On Feb 11, 5:06 pm, Antoine Deleforge 
wrote:
> Therefore, I would be very proud and glad to contribute to your
> project in anyway I can! In Particular, I noticed you participated to
> the Google Summer of Code program last 3 years, and thought it would
> be a perfect opportunity for me to get introduced to an open source
> community while combining my passions for photography and computer
> vision.

what computer system you're on? the documentation at
http://wiki.panotools.org/Development_of_Open_Source_tools is the
place to start. There is documentation for many platforms, hopefully
also the one you're using. Grab the source code, read the wiki, follow
the instructions. If something is unclear, ask questions. If you find
an improvement to the documentation, open a wiki account and edit it.
If you're usure, submit your change for discussion here. First get to
build Hugin and related tools on your computer. Then start making
small changes just for fun. Then start to make meaningful changes and
contribute a patch against the current code base. Very soon you'll be
given SVN write privilege.

For Google SUmmer of Code, past participation is no guarantee for the
future, but being already up to speed with SVN write privilege will
give you a head start over the majority of students who just drop in
at the last minute.

Good Luck
Yuv

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

[hugin-ptx] Re: 2010.1.0 testbuild (w. layout) for Win

[Yuv]

On Feb 10, 11:17 am, Jan Martin  wrote:
> No, I didn't try yet.
> I installed hugin 2009.4.0 fromhttps://launchpad.net/~jacob/+archive/ppa
> I am on Ubuntu 9.10

https://launchpad.net/~hugin is the place to go. hugin-unstable is
what you're looking for -that's trunk at revision 4964. Thanks to
Philipp Seidel for the contributions.

Yuv

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

[hugin-ptx] Re: Towards a non-patented control point detector

[Yuv]

On Feb 11, 8:26 pm, AKS-Gmail-IMAP  wrote:
> Seriously, the most obvious name is "Pablosmatic".  Everyone will know  
> exactly what it refers to.

for a short while [0]  I thought so too, but I respect Pablo's wish
for it to have another name.

if anybody has good ideas for names, post them here!

Yuv


[0] 
http://wiki.panotools.org/wiki/index.php?title=Hugin_Compiling_Ubuntu&diff=12146&oldid=12145

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

[hugin-ptx] Re: GSoC this year

[Yuv]

On Feb 9, 4:01 am, "bruno.postle"  wrote:
> I can mentor again. Though I wouldn't be useful for somebody who
> needed a lot of programming help, James basically knew exactly what he
> was doing last year and I just made suggestions.

my experience from the past three years is that most (if not all)
successful candidates did not need a lot of programming help. Daniel
had the bad luck of mentoring (and falling) those candidates that did
need programming help. AFAIK he's a lecturer in computer science. At
this stage, teaching programming is too late. Make sure there are
meaningful vetting exercises before selecting the students.

> I'd really like to see new mentors come forward, maybe some more
> previous students?

I'm happy to see James and Tim on board. Now you need a volunteer for
the admin stuff.

Yuv

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

Re: [hugin-ptx] Re: Towards a non-patented control point detector

[AKS-Gmail-IMAP]

Seriously, the most obvious name is "Pablosmatic".  Everyone will know  
exactly what it refers to.



On Feb 11, 2010, at 2:18 AM, Harry van der Wolf wrote:


Hi Pablo,

One more request. ;-)

I'm not totally happy with the name panomatic is that is already an  
existing name for an existing CP matcher/detector. Can't we call the  
panomatic binary something like "cp_matcher" or "hugin_matcher" or  
"hugin_cp_matcher" or anything else logical, not interfering and  
confusing with an existing program.

"keypoints" and "liblocalfeatures." is fine off course.

Harry


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

Re: [hugin-ptx] Output my own PTO file

[Carl von Einem]

Thanks, I added your description to


And something like 'Merge with project...' is easier to understand than
importing...

Carl

Bruno Postle schrieb am 12.02.10 00:17:
> On Thu 11-Feb-2010 at 22:43 +0100, Carl von Einem wrote:
> 
>> Although I don't know the difference between these two options, both
>> commands claim to open .pto, .ptp, .pts and .oto:
>> Menu File:
> 
>> -> Open
>> -> Import project
> 
> 'Import project' is something Thomas added to the current trunk, you can
> use it to merge another project with the current project.  If there are
> duplicate photos then only the control points are merged, otherwise new
> photos just added to the list of images.
> 
> Maybe it should be called 'Merge', since 'Import' does imply opening
> some alien file format.
> 

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

Re: [hugin-ptx] Re: Hugin's vignetting formula

[Bruno Postle]

On Thu 11-Feb-2010 at 13:45 -0800, Robert wrote:



Yes exactly.  Hugin no-longer uses a gamma correction internally to
compensate for the sensor response, it uses the EMoR model which is
an 'average' response curve with some extra parameters:


I find this a bit dubious, since the input is a camera JPEG, and
therefore I would expect the manufacturer to compensate for sensor
gain and normalise it for display purposes (i.e., apply inv. gamma).


I'm no expert, but my understanding is that digital cameras try to 
emulate the 'S' shaped response of film.



Okay, so all I have to do now is make sense of that EMoR paper by
those UOC guys. Hm, Hugin apparently uses only 5 parameters, but at
first glance in the paper and in 'emor.txt' I saw values for something
calles E(?), and curves f0-f25 (so quite a few more than just 5).


Sorry, the initial photometric model that Pablo developed proved to 
work quite well, so there has never been any pressure for user 
documentation, I suggest you look in the source code.


--
Bruno

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

Re: [hugin-ptx] Output my own PTO file

[Bruno Postle]

On Thu 11-Feb-2010 at 22:43 +0100, Carl von Einem wrote:


Although I don't know the difference between these two options, both
commands claim to open .pto, .ptp, .pts and .oto:
Menu File:



-> Open
-> Import project


'Import project' is something Thomas added to the current trunk, you 
can use it to merge another project with the current project.  If 
there are duplicate photos then only the control points are merged, 
otherwise new photos just added to the list of images.


Maybe it should be called 'Merge', since 'Import' does imply opening 
some alien file format.


--
Bruno

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

[hugin-ptx] Can I join?

[Antoine Deleforge]

Hi everybody!

Let me first introduce myself: I'm Antoine, a French student in
Computer Sciences and applied mathematics at Grenoble. I'm currently
doing a research master specialized in Computer Vision, Image
processing, Graphics and Robotics. Besides, I'm also fond of
travelling and basically can't stay in the same country more than 2
months (otherwise I get mad and some green spots start growing all
over my face!) A logical consequence of my travelling addiction is
that I usually take a lot of pictures during my trips, as an amateur,
and (yeah you guessed it) used Hugin couple of times to build some
very nice panoramas.

Therefore, I would be very proud and glad to contribute to your
project in anyway I can! In Particular, I noticed you participated to
the Google Summer of Code program last 3 years, and thought it would
be a perfect opportunity for me to get introduced to an open source
community while combining my passions for photography and computer
vision.

Hope to read from you soon,
Cheers,

Antoine.

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

[hugin-ptx] Re: Hugin's vignetting formula

[Robert]

Hi,

> >Hopefully I am overlooking something obvious... or does Hugin apply
> >some additional correction curves, even if no photometric parameter
> >except the 'a' coefficient above is set?
>
> Yes exactly.  Hugin no-longer uses a gamma correction internally to
> compensate for the sensor response, it uses the EMoR model which is
> an 'average' response curve with some extra parameters:

I find this a bit dubious, since the input is a camera JPEG, and
therefore I would expect the manufacturer to compensate for sensor
gain and normalise it for display purposes (i.e., apply inv. gamma).
So, I thought sensor gain function would be applied on top of gamma
compensation (and if all parameters are '0', only gamma should
remain-- or so I thought). However, from your response I gather that
even for EMoR parameters (0,0,0,0,0) the curve is quite different from
a simple gamma correction. In principle my engine can handle arbitrary
response curves, so it should not be impossible to compensate for
this. At least in theory.

Okay, so all I have to do now is make sense of that EMoR paper by
those UOC guys. Hm, Hugin apparently uses only 5 parameters, but at
first glance in the paper and in 'emor.txt' I saw values for something
calles E(?), and curves f0-f25 (so quite a few more than just 5). The
PanoTools wiki does not say _anything_ about the parameters (the UI
does not even name them...), apart from "you are not supposed to guess
them". I do not want to guess either, but in order to become
compatible with Hugin, I have to find out their definition...

Regards,
Robert

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

Re: [hugin-ptx] Output my own PTO file

[Carl von Einem]

Although I don't know the difference between these two options, both
commands claim to open .pto, .ptp, .pts and .oto:
Menu File:
 -> Open
 -> Import project

Jan Martin schrieb am 11.02.10 22:05:
> A crazy question:
> Can one use a PTGui .pto file with hugin?
> Or "translate" it into hugin project file?
> 

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

Re: [hugin-ptx] Re: Towards a non-patented control point detector

[Bruno Postle]

On Thu 11-Feb-2010 at 09:18 +0100, Harry van der Wolf wrote:


I'm not totally happy with the name panomatic is that is already an existing
name for an existing CP matcher/detector. Can't we call the panomatic binary
something like "cp_matcher" or "hugin_matcher" or "hugin_cp_matcher" or
anything else logical, not interfering and confusing with an existing
program.


'panomatic' is ok so long as it replaces the existing panomatic, and 
the panomatic webpage indicates this.


..but I think that ultimately this needs to be merged into Hugin and 
built as a Hugin tool, for two reasons:


1. The best interface we have for specifying parameters is a .pto 
project, this works very well for autopano-sift-c and the various 
wrappers in Panotools::Script.  So the control point generator may 
as well use the Hugin model and parser for this.


2. Feature classification really needs to be done with re-projected 
data, again the Hugin libraries are best placed to do this (the 
approach that the old match-n-shift wrapper took to create temporary 
stereographic files with PTmender had lots of problems).


--
Bruno

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

[hugin-ptx] Re: Output my own PTO file

[Jayhawk]

Thanks, Bruno. I will dive into this and report back with any other
questions I may have. Your time and help are appreciated.
Chris

On Feb 11, 3:58 pm, Bruno Postle  wrote:
> On Thu 11-Feb-2010 at 09:07 -0800, Jayhawk wrote:
>
>
>
> >I have developed some software to do some image manipulation. I am to
> >the point where I need to output my own PTO file. I have done some
> >looking around and haven't found a document which clearly states how
> >to do this (i.e., which parameters are which, required order,
> >assumptions, etc.). Could someone please point me to such a thing?
>
> The most up-to-date documentation is here, though this is a bit
> terse:http://hugin.sourceforge.net/docs/nona/nona.txt
>
> It is line-oriented, the first character of the line defines the
> data type, and a series of key/attributes follows, each separated by
> spaces.  When image numbers are referenced, counting starts at '0'.
>
> The order of lines is important, but the order of attributes within
> the lines is not.
>
> Every project needs one 'p-line' to define some attributes of the
> 'panorama' as a whole, and an 'i-line' for each input 'image'.
>
> You need 'v-lines' to drive the optimiser and 'c-lines' define a
> control point pair each.  All other lines are not so interesting.
>
> --
> Bruno

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

Re: [hugin-ptx] Hugin's vignetting formula

[Bruno Postle]

On Thu 11-Feb-2010 at 10:45 -0800, Robert wrote:


Hopefully I am overlooking something obvious... or does Hugin apply
some additional correction curves, even if no photometric parameter
except the 'a' coefficient above is set?


Yes exactly.  Hugin no-longer uses a gamma correction internally to 
compensate for the sensor response, it uses the EMoR model which is 
an 'average' response curve with some extra parameters:


http://hugin.sourceforge.net/tech/

--
Bruno

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

Re: [hugin-ptx] Output my own PTO file

[Jan Martin]

A crazy question:
Can one use a PTGui .pto file with hugin?
Or "translate" it into hugin project file?


On Thu, Feb 11, 2010 at 9:58 PM, Bruno Postle  wrote:

> On Thu 11-Feb-2010 at 09:07 -0800, Jayhawk wrote:
>
>>
>> I have developed some software to do some image manipulation. I am to
>> the point where I need to output my own PTO file. I have done some
>> looking around and haven't found a document which clearly states how
>> to do this (i.e., which parameters are which, required order,
>> assumptions, etc.). Could someone please point me to such a thing?
>>
>
> The most up-to-date documentation is here, though this is a bit terse:
> http://hugin.sourceforge.net/docs/nona/nona.txt
>
> It is line-oriented, the first character of the line defines the data type,
> and a series of key/attributes follows, each separated by spaces.  When
> image numbers are referenced, counting starts at '0'.
>
> The order of lines is important, but the order of attributes within the
> lines is not.
>
> Every project needs one 'p-line' to define some attributes of the
> 'panorama' as a whole, and an 'i-line' for each input 'image'.
>
> You need 'v-lines' to drive the optimiser and 'c-lines' define a control
> point pair each.  All other lines are not so interesting.
>
> --
> Bruno
>
>
> --
> 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 hugin-ptx@googlegroups.com
> To unsubscribe from this group, send email to
> hugin-ptx+unsubscr...@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/hugin-ptx
>



-- 
http://www.DIY-streetview.org

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

Re: [hugin-ptx] Output my own PTO file

[Bruno Postle]

On Thu 11-Feb-2010 at 09:07 -0800, Jayhawk wrote:


I have developed some software to do some image manipulation. I am to
the point where I need to output my own PTO file. I have done some
looking around and haven't found a document which clearly states how
to do this (i.e., which parameters are which, required order,
assumptions, etc.). Could someone please point me to such a thing?


The most up-to-date documentation is here, though this is a bit 
terse: http://hugin.sourceforge.net/docs/nona/nona.txt


It is line-oriented, the first character of the line defines the 
data type, and a series of key/attributes follows, each separated by 
spaces.  When image numbers are referenced, counting starts at '0'.


The order of lines is important, but the order of attributes within 
the lines is not.


Every project needs one 'p-line' to define some attributes of the 
'panorama' as a whole, and an 'i-line' for each input 'image'.


You need 'v-lines' to drive the optimiser and 'c-lines' define a 
control point pair each.  All other lines are not so interesting.


--
Bruno

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

Re: [hugin-ptx] Re: Parallax correction.

[John McAllister]

I am most grateful for the advice and support that I have received on this 
little thread, on and off list.

It appears to me that, in general, longer lenses do not demand greater/longer 
parallax adjustements than shorter lenses.
I don't want to make the adjustment range greater than it needs to be; I think 
I shall settle at about 120mm, unless somebody can cite important cases that 
exceed this factor.

The odd cases where the adjustment needs to quite small are rather challenging 
to my design; I think I shall ignore those and simply make the available 
adjustment as small as practical, about 30mm from mount.
Short adjustment also covers pocket digitals, and I have had some good results 
without being careful of PoV, at those focal lengths.

Some mad digeridoo playing hippy has pointed me at...
http://wiki.panotools.org/Entrance_Pupil_Database
Which seems pretty definitive.

I shall definitely post construction details when I have actually made this 
thing.

Thanks all...
John,
(drum playing, antique hippy: I was living on the street at 16, Jimi Hendrix 
gave me a tab of a**d, I was friends with Emily before she met Pink Floyd and I 
was the housekeeper of a Buddhist monastery 42 years ago -- and still rational!)

Love and Peace... even!

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

[hugin-ptx] Hugin's vignetting formula

[Robert]

Hi all,

I am having a bit of a problem understanding Hugin's vignetting
compensation. The background is that I am currently writing a software
library for image transformation (among others, but most importantly,
lens correction). However, while replicating the PTLens distortion
seems to be largely working (although Hugin does seem to do things a
little bit 'differently' there, too), the vignetting stuff is giving
me headaches.

Basically, as far as I understood it, the correction is supposed to be
done as a multiplicative factor c=1/(ar^6+br^4+c+1), and the corrected
channel value should be given as I = I0*c. So what I am currently
doing is the following, e.g. for the red channel:

c=1.0/(((a*r*r+b)*r*r+c)*r*r+1.0);
value_red = gamma(value_red);
value_red = value_red * c;
value_red = inv_gamma(value_red);

with the appropriate gamma correction functions, e.g. gamma(v)=v^2.2
and inv_gamma(v)=v^(1/2.2), or the equivalent sRGB gamma functions.

When I choose extreme values (e.g. c=-1), the values clip in the image
corners, which happens both in Hugin and in my implementation at the
same pixel position; so the coordinate system looks ok. Values near
the center are untouched, as they should be. However, the 'middle'
values do not quite match, no matter what I do. First I thought I was
not properly compensating for gamma, but after checking and re-
checking and double-checking I am pretty positive now that's not it...

Hopefully I am overlooking something obvious... or does Hugin apply
some additional correction curves, even if no photometric parameter
except the 'a' coefficient above is set?

Thanks and regards,
Robert

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

[hugin-ptx] Re: Parallax correction.

[Didgeridoohan]

Here you go: http://www.pannarran.se/home/panohead.html

On Feb 11, 3:45 pm, Nathan Gutman  wrote:
> Can you post a photo of the pano head that you made?

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

[hugin-ptx] Output my own PTO file

[Jayhawk]

Hello,

I have developed some software to do some image manipulation. I am to
the point where I need to output my own PTO file. I have done some
looking around and haven't found a document which clearly states how
to do this (i.e., which parameters are which, required order,
assumptions, etc.). Could someone please point me to such a thing?

For instance, I have multiple images, each with their own sift feature
locations found, the process of finding matched sift features has been
completed, and now I need to output the sift features and found
matches to a PTO file so I can do further processing with Hugin tools.

Thank you for your time and assistance.
Chris

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

Re: [hugin-ptx] Re: Parallax correction.

[Nathan Gutman]

Can you post a photo of the pano head that you made?

Didgeridoohan wrote:

Well, if you look at the Nodal Ninja heads, the small one has a scale
up to 120mm and the big one a scale up to 160mm. Myself, I made my
pano head with an allowance for up to 130mm. Should be enough for most
cases, but as Markku Kolkka said, it might be more important with the
ability to move forward to 0mm or less when it comes to longer focal
lengths.

If you look at http://wiki.panotools.org/Entrance_Pupil_Database you
see that the Sigma 70-300 at 200mm has an entrance pupil at -30mm. If
you use this on a Canon EOS 5D MKII you only have 8.5mm from the
tripod mount to the entrance pupil...

Myself, I built my pano head to be completely upgradable. I made it
with aluminium profiles (http://www.alucon.se/Shop/Image/Article/
001-001.jpg) and if needed I just make a longer arm.

  



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

[hugin-ptx] only few controlpoints were found

[Can-C . Dörtbudak]

Hi Guys,

i have a question about autopano-sift-C and panomatic. Both work vrey
stable but in my latest pics both find only few controlpoints. I have
added postit on walls to get more points but in hugin i can't find
them. I mean they were not not used and i get the error that there are
no matchting points? How do these tools work. I've tried out
autopanopro test Version and this tool gets me lot more controlpoints.
Which parameters do you use to find controlpoints in panomatic or
autopano-sift-C?

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

[hugin-ptx] Re: Parallax correction.

[Didgeridoohan]

Well, if you look at the Nodal Ninja heads, the small one has a scale
up to 120mm and the big one a scale up to 160mm. Myself, I made my
pano head with an allowance for up to 130mm. Should be enough for most
cases, but as Markku Kolkka said, it might be more important with the
ability to move forward to 0mm or less when it comes to longer focal
lengths.

If you look at http://wiki.panotools.org/Entrance_Pupil_Database you
see that the Sigma 70-300 at 200mm has an entrance pupil at -30mm. If
you use this on a Canon EOS 5D MKII you only have 8.5mm from the
tripod mount to the entrance pupil...

Myself, I built my pano head to be completely upgradable. I made it
with aluminium profiles (http://www.alucon.se/Shop/Image/Article/
001-001.jpg) and if needed I just make a longer arm.

On Feb 9, 6:03 pm, "John McAllister"  wrote:
> Evening,
>
> I am building another pano head (the fourth) which will have a rail allowing 
> the camera mount to be moved back sufficiently to eliminate parallax.
>
> I am planning to use it to produce high resolution images for very large, 
> detailed prints.
>
> Presently, I use a wide angle lens (10mm, crop 1.6=16mm 35efl) which needs 
> 100mm between the camera mount and the point of null parallax, or nodal point 
> or focal point or whatever you choose to call what I know you understand.
>
> I am aiming to move out to focal lengths of perhaps 100mm, or some more.
>
> Does anybody have any particular experience with longer focal lengths, 
> particularly with respect to zoom telephoto lenses (haven't bought one yet, 
> so can't experiment)?
>
> I want to determine a sensible maximum allowance for parallax elimination 
> with longer and zoom lenses.
>
> Could anybody enlighten me? (Buddhists tried)
>
> Ever so humbly grateful in advance of what I know will be a deluge of 
> erudition and flawless advice.
>
> Brainache (not Brainiac, obscure reference there folks).
>
> PS. See some lucky pictures of my garden foxes (short slideshow) 
> athttp://www.w3a2z.net/Rasha/

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

Re: [hugin-ptx] 2010.1.0 testbuild (w. layout) for Win

[Oskar Sander]

Cheers mate

2010/2/10 brian_ims 

>
> Oskar
>
> Search for General Panini by Tom Sharpless
>
> Brian
>
>
> Oskar Sander wrote:
> >
> > Pardon my ignorance, but I though I saw recently someone posting a win
> > test-build on the list. However, I can't find it searching.  Is there one
> > out there, or did I just mix it up with the OSX-version?
> >
> > Cheers
> > /O
> >
> > --
> > 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 hugin-ptx@googlegroups.com
> > To unsubscribe from this group, send email to
> > hugin-ptx+unsubscr...@googlegroups.com
> > For more options, visit this group at
> > http://groups.google.com/group/hugin-ptx
> >
>
> --
> View this message in context:
> http://old.nabble.com/2010.1.0-testbuild-%28w.-layout%29-for-Win-tp27533386p27538800.html
> Sent from the hugin ptx mailing list archive at Nabble.com.
>
> --
> 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 hugin-ptx@googlegroups.com
> To unsubscribe from this group, send email to
> hugin-ptx+unsubscr...@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/hugin-ptx
>



-- 
/O

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

Re: [hugin-ptx] Pano head construction and parallax elimination.

[John McAllister]

By deep, I meant a large VFoV.
I'm rather hoping to undertake some interior survey work.

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

[hugin-ptx] Re: Enblend artifact problem

[distudio]

On Feb 11, 5:30 pm, cspiel  wrote:
> Rob -
>
> On Feb 11, 3:25 am, distudio  wrote:
>
> > I thought I might have stumbled upon a solution. I ran a problem
> > script with the -a switch in the enblend command line, no go.
>
> It looks like SF bug ID 2863465 to me.

Indeed it does, however in my case the panos generally cover under 180
degree HOAV plus I'm using Enblend 4 not 3.2 as per bug ID 2863465.
Plus the problem projects are generally single row left to right
overlap.

Does anyone know what a Priority: 6 bug is? I'd hope this is something
that the developers might consider a priority as it's going to become
a significant issue as pano sizes creep up.

I tried to use Smartblend as a substitute blender but it's broken for
use with the current version of Hugin and the batch patch that was
produced doesn't appear to work either. Not an ideal situation.

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

Re: [hugin-ptx] Re: Towards a non-patented control point detector

[Harry van der Wolf]

Hi Pablo,

One more request. ;-)

I'm not totally happy with the name panomatic is that is already an existing
name for an existing CP matcher/detector. Can't we call the panomatic binary
something like "cp_matcher" or "hugin_matcher" or "hugin_cp_matcher" or
anything else logical, not interfering and confusing with an existing
program.
"keypoints" and "liblocalfeatures." is fine off course.

Harry


2010/2/11 Harry van der Wolf 

> Hello Pablo,
>
> Thanks a lot! Has anyone ever mentioned that you are important to this Open
> source project? :-)
>
> 2010/2/11 Pablo d'Angelo 
>
> Hi Harry,
>>
>>
>> Harry van der Wolf wrote:
>>
>>  - The mail mentions that a "a small part of the SURF algorithm to
>>> estimate the orientation of the interest point" is still used. You mention
>>> that you hope to replace this in the future. Can you give a rough estimate
>>> or could this be a GSOC 2010 project?
>>>
>>
>> Actually, I have already implemented another orientation estimation
>> algorithm, which seems to work quite well.
>>
>>
>>
> Shame on me: I didn't check the repository lately. I will start testing
> asap.
>
> Does this means that "we" now have a fully patent-free CP detector and one
> that works with Thomas' multi-step CPdetector setup as well? The latter due
> to the fact that we can use the keypoints binary (also patent-free I
> assume?) for these more complicated setups?
>
> If so, it means that "we" can now release Hugin on all platforms with a CP
> detector without having to be affraid of any patent/license issues.
>
> Hoi,
> Harry
>
>
>

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