[Insight-developers] Change to itkMultiResolutionPyramidImageFilter.txx
Hans Johnson
hans-johnson at uiowa.edu
Thu Mar 26 17:43:38 EDT 2009
Brad,
The problem is that the origin is not properly computed in the down sampled
images, and the origin affects the location.
Hans
--
Hans J. Johnson, Ph.D.
Hans-johnson at uiowa.edu
278 GH
The University of Iowa
Iowa City, IA 52241
(319) 353 8587
From: Bradley Lowekamp <blowekamp at mail.nih.gov>
Date: Thu, 26 Mar 2009 11:46:03 -0400
To: Hans Johnson <hans-johnson at uiowa.edu>
Cc: Bill Lorensen <bill.lorensen at gmail.com>, Luis Ibanez
<luis.ibanez at kitware.com>, ITK <insight-developers at itk.org>
Subject: Re: [Insight-developers] Change to
itkMultiResolutionPyramidImageFilter.txx
Hello Hans,
I don't see any definition for center of mass there. A discrete center of
mass is usually defined as a sum ( weight * location)/ sum( weight ). I do
not know if what you are intending weight to mean in your definition. I have
been assuming that you are not referring to any, and all the weights are
one. Therefore, better terminology may be "physical center" or "physical
location of center". Which are you referring too?
If this was just a bug with the geometry, the Spacing and Origin meta data,
then why were there changes to the sampling algorithms? I think that the
geometry needs to be clearly explained some place.
In terms of a diagram I think that we need something equivalent to the
Figure 4.1 in the Guide, but for resampling (perhaps only 1d). Are we
assuming that there is a 1-to-1 mapping of all physical points between an
original and a resampled image? This is where I see the heart of this
problem.
I don't see how changing from nearest to linear interpolation is related. It
seems more like change in behavior and output then this bug (the way I
understand it).
Also if we look at the changes to Shrinker:
http://public.kitware.com/cgi-bin/viewcvs.cgi/Code/BasicFilters/itkShrinkIma
geFilter.txx?root=Insight&r1=1.54&r2=1.55&sortby=date
I have not looked at the itkShrinkImageFilter::GenerateOutputInformation
method close enough to know what it's trying to do. I find the addition of
the transform to be rather poor technique, and a reduction in the light
weight performance this class previously had.
But the question still remains is changing from the Shrinker to a linear
interpolated Resample filter really a bug? or is it a fundamental change in
behavior?
Brad
On Mar 26, 2009, at 11:03 AM, Hans Johnson wrote:
> Brad,
>
> Look at the attachment to bug http://public.kitware.com/Bug/view.php?id=8275.
> This attachment includes the image and the description of center of mass bug.
>
> Hans
> --
> Hans J. Johnson, Ph.D.
> Hans-johnson at uiowa.edu
>
> 278 GH
> The University of Iowa
> Iowa City, IA 52241
> (319) 353 8587
>
>
>
>
> From: Bradley Lowekamp <blowekamp at mail.nih.gov>
> Date: Thu, 26 Mar 2009 10:57:23 -0400
> To: Bill Lorensen <bill.lorensen at gmail.com>
> Cc: Luis Ibanez <luis.ibanez at kitware.com>, ITK <insight-developers at itk.org>,
> Hans Johnson <hans-johnson at uiowa.edu>
> Subject: Re: [Insight-developers] Change to
> itkMultiResolutionPyramidImageFilter.txx
>
> I am giving these changes a closer look. I don' think they were the correct
> thing to do as they made a fundamental change in the behavior of the Pyramids.
>
>
> http://www.itk.org/cgi-bin/viewcvs.cgi/Code/Algorithms/itkMultiResolutionPyram
> idImageFilter.txx?root=Insight&r1=1.25&r2=1.26&sortby=date
>
> Changing from the Shrinker (which used to be an optimized nearest neighbor)
> to using a linear interpolated Resample seems wrong to me.
>
> I don't fully understand what the "center of mass" bug is. I think a more
> documentation and a diagram is really needed, to describe the behavior of
> resampling, and shrinking. This will prevent another bug from happening again
> when the current behavior is not liked.
>
> Brad
>
> ps
> I could talk more about optimization of itkShrinkImageFilter has been reduced
> too, but I think that is another subject.
>
>
> On Mar 25, 2009, at 4:33 PM, Bill Lorensen wrote:
>
>
>> Luis,
>>
>> This change
>>
>>
>> http://www.itk.org/cgi-bin/viewcvs.cgi/Code/Algorithms/itkMultiResolutionPyra
>> midImageFilter.txx?root=Insight&r1=1.29&r2=1.30&sortby=date
>>
>> should be added to the 3.10 release. This undoes the backward
>> compatibility I broke with a bad checkin on Dec 17
>>
>> http://www.itk.org/cgi-bin/viewcvs.cgi/Code/Algorithms/itkMultiResolutionPyra
>> midImageFilter.txx?root=Insight&r1=1.26&r2=1.27&sortby=date
>>
>> Bill
>>
>>
>> Notice: This UI Health Care e-mail (including attachments) is covered by the
>> Electronic Communications Privacy Act, 18 U.S.C. 2510-2521, is confidential
>> and may be legally privileged. If you are not the intended recipient, you
>> are hereby notified that any retention, dissemination, distribution, or
>> copying of this communication is strictly prohibited. Please reply to the
>> sender that you have received the message in error, then delete it. Thank
>> you.
>>
>> _______________________________________________
>> Powered by www.kitware.com <http://www.kitware.com>
>>
>> Visit other Kitware open-source projects at
>> http://www.kitware.com/opensource/opensource.html
>>
>> Please keep messages on-topic and check the ITK FAQ at:
>> http://www.itk.org/Wiki/ITK_FAQ
>>
>> Follow this link to subscribe/unsubscribe:
>> http://www.itk.org/mailman/listinfo/insight-developers
>>
>
>
> ========================================================
>
> Bradley Lowekamp
>
> Lockheed Martin Contractor for
>
> Office of High Performance Computing and Communications
>
> National Library of Medicine
>
> blowekamp at mail.nih.gov
>
>
>
>
>
>
========================================================
Bradley Lowekamp
Lockheed Martin Contractor for
Office of High Performance Computing and Communications
National Library of Medicine
blowekamp at mail.nih.gov
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/mailman/private/insight-developers/attachments/20090326/59395cd8/attachment.htm>
More information about the Insight-developers
mailing list