[Insight-developers] ObjectMorphologyImageFilterTest

[Joshua Cates]

Hi Stephen,

> Actually, it looks like there is a flaw in my logic in handling faces. 

Not necessarily, I would probably use the same approach unless kernel
sizes are dramatically different.  But there will be a performance penalty
on the iterator using the smaller kernel at the boundary faces.  An
iterator will always do bounds checking if defined so that its region
touches the image boundary. So you will lose some efficiency with
increasing difference in the two kernel sizes.

But I wouldn't expect to see a dramatic slowdown because of this unless
you are processing a data set with a high surface area to volume ratio.

We can always add a method to manually turn off bounds checking if that 
would be helpful.  Then your algorithm has the option of deciding when to 
turn it on and off for a particular iterator.

Josh.

______________________________
 Josh Cates			
 School of Computer Science	
 University of Utah
 Email: cates@sci.utah.edu
 Phone: (801) 587-7697
 URL:   http://www.sci.utah.edu/~cates


On Thu, 9 Jan 2003, Stephen R. Aylward wrote:

> Hi Josh,
> 
> It is my fault - I think you and I checked stuff in on nearly the same day.
> 
> Actually, it looks like there is a flaw in my logic in handling faces. 
>   I am using two neighborhood iterators and have the one with the larger 
> kernel size drive the definition of the faces, but it looks like I will 
> have to calculate the faces for each iterator if their kernels are 
> different sizes.   Is that correct?
> 
> s
> 
> 
> Joshua Cates wrote:
> > Hi Stephen, Bill,
> > 
> > Let me know what I can do to help.  I thought I had found and fixed all 
> > the appropriate files, but I guess I missed some.  The only change needed 
> > to any filter should be to replace SmartNeighborhoodIterator and 
> > RandomAccessNeighborhoodIterator with NeighborhoodIterator.
> > 
> > Josh.
> > 
> > ______________________________
> >  Josh Cates			
> >  School of Computer Science	
> >  University of Utah
> >  Email: cates@sci.utah.edu
> >  Phone: (801) 587-7697
> >  URL:   http://www.sci.utah.edu/~cates
> > 
> > 
> > On Thu, 9 Jan 2003, Lorensen, William E (Research) wrote:
> > 
> > 
> >>Stephen,
> >>	Could you take a shot at converting the Object morphology filters over to the new
> >>NeighborhoodIterator API? I'm sure Josh can give guidance if you run into trouble. Currently, it
> >>spews dozens of deprecation messages and sometimes hangs on the continuous builds.
> >>
> >>Thanks,
> >>
> >>Bill
> >>
> >>_______________________________________________
> >>Insight-developers mailing list
> >>Insight-developers@public.kitware.com
> >>http://public.kitware.com/mailman/listinfo/insight-developers
> >>
> > 
> > 
> > _______________________________________________
> > Insight-developers mailing list
> > Insight-developers@public.kitware.com
> > http://public.kitware.com/mailman/listinfo/insight-developers
> 
> 
> -- 
> ===============================================
> Dr. Stephen R. Aylward
> Assistant Professor of Radiology
> Adjunct Assistant Professor of Computer Science
> http://caddlab.rad.unc.edu
> aylward@unc.edu
> (919) 966-9695
> 
>