[Insight-developers] Neighborhoods and filter output types

Thu, 18 Oct 2001 08:38:01 -0400

This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

------_=_NextPart_001_01C157D1.B9358B80
Content-Type: text/plain

I'm not sure that the right thing to do here.  One the one hand,  using the chain rule is should be
pretty easy to convert the derivative calculated in parametric (index) space into a derivative in
physical space.  I imagine it is actually easier to do this rather than build the derivative kernel
to account for the spacing.  Filters like the DerivativeImageFilter, GradientImageFilter, etc could
do this conversion automatically and leave the DerivativeOperator as calculating the derivative in
parametric space. On the other hand, will people "forget" to do this conversion.

I imagine the same issue of spacing exists in the GaussianOperator (and hence the
DiscreteGaussianImageFilter).

-----Original Message-----
From: Joshua Cates [ mailto:cates@cs.utah.edu <mailto:cates@cs.utah.edu> ]
Sent: Wednesday, October 17, 2001 2:21 PM
To: Miller, James V (CRD)
Cc: Insight-developers (E-mail)
Subject: Re: [Insight-developers] Neighborhoods and filter output types

Hi Jim,

> 
> DerivativeOperator
> 
> Should the derivative operator take into account the spacing of the data?  Or should the derivative
> just be in parametric space and whomever uses the derivative operator (like the
> DerivativeImageFilter) must decide to convert the values to physical space.  I can go either way on
> this one.
> 

Ross and I have talked about making the conversion to use the spacing of
the data.  I won't have time to do this any time in the near future, but
if people generally agree this is a good idea, then I'll get to it
eventually.

Josh.

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

------_=_NextPart_001_01C157D1.B9358B80
Content-Type: text/html

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=us-ascii">
<TITLE></TITLE>

<META content="MSHTML 5.50.4807.2300" name=GENERATOR></HEAD>
<BODY>
I'm not sure that the right thing to do here.&nbsp; One the one 
hand,&nbsp;&nbsp;using the chain rule is should be pretty easy to convert the 
derivative calculated in parametric (index) space into a derivative in physical 
space.&nbsp;&nbsp;I imagine it is actually easier to&nbsp;do this&nbsp;rather 
than build the derivative kernel to account for the spacing.&nbsp; Filters like 
the DerivativeImageFilter, GradientImageFilter, etc could&nbsp;do this 
conversion automatically and leave the&nbsp;DerivativeOperator as calculating 
the derivative in parametric space. On the other hand, will people "forget" to 
do this conversion.
I imagine the same issue of spacing exists in the 
GaussianOperator (and hence the DiscreteGaussianImageFilter).
 -----Original 
Message----- From: Joshua Cates [<A 
href="mailto:cates@cs.utah.edu">mailto:cates@cs.utah.edu</A>] Sent: 
Wednesday, October 17, 2001 2:21 PM To: Miller, James V (CRD) Cc: 
Insight-developers (E-mail) Subject: Re: [Insight-developers] Neighborhoods 
and filter output types Hi Jim, &gt;&nbsp; &gt; 
DerivativeOperator &gt;&nbsp; &gt; Should the derivative operator take 
into account the spacing of the data?&nbsp; Or should the derivative &gt; 
just be in parametric space and whomever uses the derivative operator (like 
the &gt; DerivativeImageFilter) must decide to convert the values to physical 
space.&nbsp; I can go either way on &gt; this one. &gt;&nbsp; Ross 
and I have talked about making the conversion to use the spacing of the 
data.&nbsp; I won't have time to do this any time in the near future, but if 
people generally agree this is a good idea, then I'll get to 
it eventually. Josh. ____________________________ &nbsp;Josh 
Cates&nbsp;&nbsp;&nbsp;&nbsp; &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; 
&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; &nbsp;School of Computer 
Science&nbsp;&nbsp;&nbsp;&nbsp; &nbsp;University of Utah &nbsp;Email: 
cates@cs.utah.edu &nbsp;Phone: (801) 587-7697 &nbsp;URL:&nbsp;&nbsp; 
www.cs.utk.edu/~cates </BODY></HTML>

------_=_NextPart_001_01C157D1.B9358B80--