[Insight-developers] Blox Images

George Stetten stetten@pitt.edu
Wed, 30 Jan 2002 09:39:23 -0500


The trouble is that creating a new image will be very expensive, even if we are removing
or adding CoreAtomPixels.  Is there any way that a filter can have the same image as its
input and output?

"Miller, James V (CRD)" wrote:

> > The remaining operations one performs on a core atom image, eigenanalysis
> > and voting, modify ivars of CoreAtomPixels - you don't get a "new" image as
> > a result, so I'm uncomfortable implementing these as filters. If you did,
> > you'd either:
>
> If CoreAtomPixels are "pixels" in a BloxImage, then an operation that modifies
> ivars of CoreAtomPixels are creating a "new" image (the output CoreAtomPixels are
> different from the input CoreAtomPixels).
>
> I don't know enough about the "voting" operation and "eigenanalysis" operations
> on BloxImages to say whether these operations produce "new" images or not (and
> hence whether they are filters or not).
>
> But if "voting" reduces the number of CoreAtoms in a BloxImage by removing
> inconsistent atoms, then it does produce a new image.
>
> If "eigenanalysis" creates a series of "measurements" from a BloxImage, then again
> I would say it is a filter.  It may not produce another BloxImage, it may not produce
> an image at all.  But it does take an input image and calculate (or recalculate)
> a series of measurements.  Maybe these operations are more "calculators" but I would
> push for people to think of these operations as "filters".
>
> _______________________________________________
> Insight-developers mailing list
> Insight-developers@public.kitware.com
> http://public.kitware.com/mailman/listinfo/insight-developers

--
George Stetten, M.D., Ph.D.
Assistant Professor, Bioengineering, U. Pitt.
Research Scientist, Robotics Institute, CMU
http://www.stetten.com