[Insight-users] Image Piece Overlap Registration

[Luis Ibanez]

Hi Kevin,

If you can count on the microscope to perform translation without
any rotations of the plate, then you can restrict your registration
to use a TranslationTransform.  Since your images are of the same
modality you probably should try the NormalizedCorrelation image Metric
first.

This metric will work well with a RegularStep gradient descent
algorithm.


VERY IMPORTANT: Please make sure that you set the pixel size (spacing)
in your images correctly, as well as the origins of the images. If you
don't have reliable information about the origins, then you can set them
to zero and rather use your knowledge about the approximate translation
of the microscope in order to initialize the Transform before starting
the registration.

Note also that since your images will have a very thin overlap, it is
very important that you set the "FixedImageRegion" to cover only the
section of the fixed image that you expect will be overlapping the
moving image.


We will be interested to hear about your findings. BTW that will be
a nice submission to the Insight Journal.



    Regards,


        Luis



-----------------------
Kevin H. Hobbs wrote:
> My coworker scanned many ganglia in our confocal microscope. He needed
> to use a very hi resolution which limits the working distance. In order
> to get all of each cell he had to scan them in several pieces moving the
> microscope stage for each piece. 
> 
> My task is now to reassemble these pieces and start doing some analysis
> but unfortunately when the stage is moved the origin for each piece is
> not changed. The first thing I have to do is register the pieces with
> one another and stitch them together.
> 
> I don't think this should be too hard but if anybody has any suggestions
> I'd like to hear them, and if anybody is interested I'll share what I
> find out.
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users