[Insight-users] Possible bug in itkAnalyzeImageIO?

[Raghavendra Chandrashekara]

Hi Kent,

I think I am using the 1.2 release. This is what I typed in to update our
CVS repository:

cvs update -r Release-1-2-0 -d -P

Is that correct? When I explicitly set the IO object to be the GIPL image
reader then everything works fine.

Raghavendra.

----- Original Message -----
From: "Kent Williams" <kent@mail.psychiatry.uiowa.edu>
To: "Raghavendra Chandrashekara" <rc3@doc.ic.ac.uk>
Sent: Wednesday, March 19, 2003 5:27 PM
Subject: Re: [Insight-users] Possible bug in itkAnalyzeImageIO?


The Analyze IO class is a subject of current development; the current
version
in CVS does not have this undesirable behavior.  I don't know what version
of
ITK you're working with, but the 1.2 release doesn't do this.

If you've patched Code/IO/itkImageIOFactory.cxx to register the Analyze
filter
in the ITK 1.1 code, you are guaranteed to have a problem, because the
Analzye Image IO had exactly the problem you're describing.

If for some reason you must stay with ITK 1.1, you could try checking out
itkAnalyzeImageIO.cxx from cvs with the ITK 1.2 tag. That version is more
robust, and should coexist with the 1.1 ITK tree.

I'm glad to assist you in any way I can with this if you have further
questions.

On Wednesday 19 March 2003 11:08 am, Raghavendra Chandrashekara wrote:
> Hi all,
>
> I think that there might be another bug in the itkAnalyzeImageIO class.
> It occurs when reading GIPL images. It looks like an exception:
>
> itk::ExceptionObject (0x815bc38)
> Location: "Unknown"
> File: /vol/vipdata/packages/InsightCVS/Code/IO/itkAnalyzeImageIO.cxx
> Line: 1002
> Description: File cannot be read
>
> is thrown before the GIPL IO class has had a chance to look at the
> image. I have attached the image causing the problems.
>
> Thanks,
>
> Raghavendra.