Login  Register

Re: Allignement of images using correlate (FD math)

Posted by Saalfeld, Stephan on Feb 22, 2014; 1:12pm
URL: http://imagej.273.s1.nabble.com/Allignement-of-images-using-correlate-FD-math-tp5006622p5006630.html

Hi Philippe,

to answer your question more elaborately:  The return image of FD
Math>Correlate does not contain information about rotation or tilt.  You
cannot correct for rotation and tilt with the return image of FD
Math>Correlate.  Do you mean shear when you say tilt?.

If you like to stick with FD, there is a not so robust possibility to
identify rotation.  Polar-transform the FTs of both images and then do
FD Math>Correlate to identify their translation.  This translation
corresponds to the angle of rotation, not the translation.

For efficient information about writing the lines in your own code study
the code and references of the mentioned plugins that all implement
efficient solutions to this problem.

Cheers,
Stephan


On Sat, 2014-02-22 at 11:38 +0100, Philippe GENDRE wrote:

> Dear Stephan and the whole List,
>
> Thank for your answer Stephan.
>
> I am not looking for a plugin to do that but efficient information
> according to write the lines in my own code (for different reasons).
>
> For example I found this :
> http://www.u-bordeaux1.fr/crazybiocomputing/docs/tbs225/cours_5.pdf for
> translation correction (see page 8-10) which helps me a lot.
>
> But how to correct for rotation or tilt mismatch ?
>
> Best regards,
>
> Philippe
>
>
> 2014-02-22 0:15 GMT+01:00 Stephan Saalfeld <[hidden email]>:
>
> > You cannot.  Check the plugins
> >
> > http://fiji.sc/StackReg
> >
> > http://fiji.sc/Linear_Stack_Alignment_with_SIFT
> >
> > http://fiji.sc/Register_Virtual_Stack_Slices
> >
> > http://fiji.sc/BUnwarpJ
> >
> > or
> >
> > http://fiji.sc/TrakEM2
> >
> > Cheers,
> > Stephan
> >
> >
> >
> >
> >
> >
> >
> > On Fri, 2014-02-21 at 22:52 +0100, Philippe GENDRE wrote:
> > > Dear List,
> > >
> > > I am currently using FD math --> Correlate to allign a set of images
> > from a
> > > reference.
> > >
> > > I am using the position of the maximum in the "result" image according to
> > > apply a translation operation on the set of images.
> > >
> > > Hqw to correct my set of images from rotation or tilt using the
> > information
> > > in the "result" image ?
> > >
> > > Best regards,
> > >
> > > Philippe
> > >
> > > --
> > > ImageJ mailing list: http://imagej.nih.gov/ij/list.html
> >
> > --
> > ImageJ mailing list: http://imagej.nih.gov/ij/list.html
> >
>
> --
> ImageJ mailing list: http://imagej.nih.gov/ij/list.html

--

Stephan Saalfeld, Ph.D.
Lab Head
Janelia Farm Research Campus
19700 Helix Drive | Ashburn, VA 20147
Phone: 571-209-4184 | Fax: 571-209-4946
[hidden email]

--
ImageJ mailing list: http://imagej.nih.gov/ij/list.html