Colour/brightness correction (was:hugin update)

[Kai-Uwe Behrmann]

Hi Alexandre,

Am 11.11.03, 09:29 +0100 schrieb alexandre jenny:

> Because it works this way :
>  - from the ancor image, let's find the neightboors.
>  - do
>  -   first neightboor, do an histogram adjustement with 256 variables.

Does this mean the adjustment is done despite of any position related
differences in the overlapping zone?  ...hmm ..
Lets imaging an sunset example for an full pano.
Image A with sun in the right part and much lens flare. The following
image B on the right side of image A does not have as much flare but a
little. The top image C covers a part of bright sky nearly no flare, the
lens is not hit by the sun. ...some more images to complete the pano.
Now I think it should be possible to apply an transition to every image
which solves the lens flare and correct the images region specific.
Clearly image C would become an anchor. The other images should then
adjusted according to the light differences on the right overlap region.
And on the left overlaping region, according to the differences there. To
do this stepless an transition from the right correction value to the
left one would be best to make the correction invisible. The scenario is
somewhat specific, but I hope this could make understand what kind of
color/brightness correction I need.

I sit on this and the most problem for me is at the moment to understand
the math to work with bezier splines. My approach is to build a mesh of
wighted middle lines of the overlapping regions. This mesh needs to be
really middle and identical for each image. At as many positions as needed
can an color difference been measured by approximating a number of
pixels and then checked against too high detail frequencys (ghosting
people) ... The resulting correction value I want to apply for each pixel
in the image separately depending from its distance to the two next
measurement points and the image centre.
The spline is needed to find the middle line. All other things I can
clearly imagin.

Now can Your approach solve such an scenario? So maybe my approach is too
complicated and the flare issue is solved otherwise.


> I am actually working in a feature extration tools that could be used
> with hugin. Every of you I think have read this
> http://www.cs.ubc.ca/~mbrown/panorama/panorama.html. That's exactly what
> I'm rewritting. I have already the SIFT feacture extractor. I just need

Sounds good.

> some more routines to be able to find pair of picture and place them
> approximately. My option is not to do a full panorama tools from picture
> to pano, but to generate a hugin or ptgui project directly with the good
> parameters. I called this project PtPrepros (panoTools preprocessor) and
> it should be available in a week or two. The only big issue I have is a
> licencing problem. I have to ask Dr Lowe for that.

Good luck with this question.

regards
Kai-Uwe