[Insight-developers] ITK: A successful implementation story

[Hans Johnson]

Hello All,

Today has been a very productive day, and I wanted to share my appreciation
to the entire ITK group for making this possible.  This is just one of many
examples I could give on how the ITK community has accelerated and improved
our local research efforts.

My graduate student has been doing some atlas based segmentation were signed
distance maps of binary objects are passively deformed.  One of the major
bottle necks is creating the signed distance maps with the Danielson
Distance Map, and the Future work section of his thesis stated that the work
by Maurer should be investigated as a possible way to speed this up.

Timeline to success:

9:26am  e-mail arrives about a new Insight Journal submission
12:30am print and read documentation while
12:45am download source code, replace Dannielson Filter with new EDT filter,
and compile
 1:00pm run on 1 dataset
 1:03pm curse because it could not possibly have been done in only 3
minutes.
 1:10pm compare new results to old results, do dance of joy because new
results are nearly the same as old results
 4:00pm Write Insight Journal review

Good work Penn Image Computing and Science Laboratory!

Thanks,
Hans

------ Forwarded Message
From: <vjammala at engineering.uiowa.edu>
Date: Fri, 17 Feb 2006 13:42:17 -0600
To: Hans Johnson <hans-johnson at uiowa.edu>
Subject: Re: [Fwd: The Insight Journal-New Submission]

Dr Johnson,
             There is a ten fold improvement in the speed using the new
filter.
Previously to generate a signed distance mask it used to take about 40
minutes
and now it is done in  3 minutes. I have warped the new signed distance
masks
to generate output masks , the output masks are similar to the old result
but
not exactly the same.
Thanks
vamsi

Quoting Hans Johnson <hans-johnson at uiowa.edu>:

> Vamsi,
>
> Download, and implement this filter as FastSignedDistance with the
> same command line parameters as SignedDistance.  It is important that
> you implement this as quickly as possible, and keep track of how long
> it takes.
>
> The reason for implementing this function is to describe how great
> ITK is.  You can incorporate other peoples work and verify that it
> does what it is supposed to in very little time.
>
> Hans
>
>

------ End of Forwarded Message
------ Forwarded Message
From: <webmaster at insightsoftwareconsortium.org>
Reply-To: <webmaster at insightsoftwareconsortium.org>
Date: Fri, 17 Feb 2006 10:26:07 -0500
To: <hans-johnson at uiowa.edu>
Subject: The Insight Journal-New Submission

Hello,
A new submission has been added to the Insight Journal.

Title: N-D Linear Time Exact Signed Euclidean Distance Transform
Author(s): Tustison, Nicholas J.; Siqueira, Marcelo; Gee, James C.
Abstract: 
Fast computation of distance transforms find direct application in various
computer vision problems.  Currently there exists two image filters in the
ITK library which can be used to generate distance maps.  Unfortunately,
these filters produce only approximations to the Euclidean Distance
Transform (EDT).  We introduce into the ITK library a third EDT filter which
was developed by Maurer {em et al.} cite{Maurer2003}. In contrast to other
algorithms, this algorithm produces the exact signed squared EDT using
integer arithmetic.  The complexity, which is formally verified, is $O(n)$
with a small time constant where $n$ is the number of image pixels.

Download the paper at: http://hdl.handle.net/1926/171
Review the paper at: http://insight-journal.org

Generated by the Insight Journal Manager

You are receiving this email because you asked to be informed by the Insight
Journal for new submissions.
To change your email preference go to http://insight-journal.org.

------ End of Forwarded Message