[Insight-users] Problem with SignedMaurerDistanceMapImageFilter

[Andriy Fedorov]

Hi, Nick,

Thanks a lot for your reply.

No I have not figured this out.

My assumption is, that I should be able to get all the limitations and
all the instructions on how to use the filter from the Doxygen. I cite
it here: "Set/GetBackgroundValue specifies the background [...], other
than that the usage is completely analogous to the
itkDanielssonDistanceImageFilterClass except is does not return the
Voronoi map [...]"
(http://www.itk.org/Doxygen/html/classSignedMaurerDistanceMapImageFilter.html)
Danielsson works just fine for the same data where Maurer fails!

I did not spend a lot of time trying to figure it out, because I
thought that whenever the filter behavior is different from whatever
is written in the documentation - this should be a bug.

My opinion is that either Maurer filter should be updated so that it
doesn't fail for arbitrary input OR the documentation should be
updated to reflect its limitations.

I hope ITK developers can add more clarity to this issue.

Thanks,
Andriy Fedorov



On 12 Oct 2007 10:27:55 -0400, Nick Tustison <ntustison at wustl.edu> wrote:
> Hi Andriy,
>
> I was not assigned the bug associated with your problem but as the
> contributor of the filter, I thought I'd comment on the difficulties you are
> encountering.  In case you haven't figured it out, the Maurer distance
> transform does indeed fail for inputs in which a foreground pixel is on the
> image boundary.  When I wrote the filter, I thought about simply passing the
> input image through an instantiation of the ConstantPadImageFilter, creating
> the distance image on the padded image and then extracting the original
> region but ultimately decided against it for reasons I can no longer
> remember.  Perhaps Luis, or others, can weigh in as to where the padding
> should be performed if there are foreground pixels on the boundary.  In the
> interim, it is an easy work around.
>
> Good luck,
> Nick Tustison
>
>
>
>
>
>
>
>
> On 10/11/07, Andriy Fedorov <fedorov at bwh.harvard.edu> wrote:
> >
> > Hello,
> >
> > I observed that the Maurer distance filter fails for some inputs. I
> > submitted the bug and posted an example image on June 15 here
> > http://www.itk.org/Bug/view.php?id=5198 , but I don't see any
> > development of this issue, and the status remains "assigned".
> >
> > I am curious, what is the usual time frame and schedule for resolution
> > of critical and reproducible bugs in the release version of ITK? Not
> > that it is extremely critical to me that this bug is fixed, but I am
> > surprised the issue remains open without attention for so long.
> >
> > Thanks,
> >
> > Andriy Fedorov
> > _______________________________________________
> > Insight-users mailing list
> > Insight-users at itk.org
> > http://www.itk.org/mailman/listinfo/insight-users
> >
>
>