Text archives Help
- From: Jonathan Bronson <bronson@sci.utah.edu>
- To: Johan Andruejol <johan.andruejol@kitware.com>
- Cc: cleaver <cleaver@sci.utah.edu>
- Subject: [Cleaver] Re: Re: Re: Fwd: Cleaver fails on labelmap
- Date: Thu, 05 Dec 2013 16:04:15 -0700
I've verified that it is an
ill-conditioned matrix at fault.
One or more of the 4-material interfaces is poorly
defined (determinant = 0) by the provided input fields.
If this happens while computing 3-material interfaces, the
code falls back to using a barycenter as the location. (Those
are the other warnings you're getting) Currently, the 4-material
interface code has no such fallback case.
I'll go ahead and produce a patch for this and send out an
email when its live.
-Jonathan
On 12/05/2013 03:38 PM, Jonathan Bronson wrote:
52A10049.4050005@sci.utah.edu" type="cite">
Ah, it looks like I left off the 4th
material.
I can now reproduce the problem. I'll look
into this in more detail and get back to you.
-Jonathan
On 12/05/2013 03:31 PM, Johan Andruejol wrote:
CA+LOGz1=QeBLmUWjJ2U2-jFDF8gB8T_iQEsDHCjx43WWiU8nWA@mail.gmail.com"
type="cite">
Hi Jonathan,
Thanks for the quick response.
I don't use any special kind of parameters. I just run the
command: cleaver -p -i image0.nrrd image1.nrrd image2.nrrd
image3.nrrd -o output.
I use cleaver 1.5.3.
Johan
|
Archive powered by MHonArc 2.6.16.