[Insight-developers] ImageIO & supported Pixel types -- TMP programming challenge

[wanlin]

It's great. I am thinking whether it is better input a pixel instance or a
pixel pointer point to the first pixel in image instead of null, so it is
cable get length for Array or VariableLengthVector to support writing them
as well.



On Sat, Nov 13, 2010 at 12:32 AM, kent williams <norman-k-williams at uiowa.edu
> wrote:

>  Thanks! I actually figured this out last night — a solution almost
> identical to yours. Only difference is that instead of a pixel instance
> type, e.g.
>
> SetPixelTypeInfo(const TPixel **) { /** set pixel stuff */ }
>
> And then call it
>
> SetPixelTypeInfo(static_cast<const TPixel *>(0));
>
>
> On 11/12/10 2:13 AM, "wanlin" <wanlinzhu at gmail.com> wrote:
>
> HI, Kent,
>          I think one possible way to handle this issue is to determine
> pixel component type and pixel type by input a PixelType value in
> ImageIOBase::SetPixelTypeInfo() but not type_info.
> I upload 4 patches to demonstrate how could it work.
>
>
>
>
>
>
>
> kind regards
>
>
>
> wanlin
>
>
>
> On Fri, Nov 12, 2010 at 6:28 AM, kent williams <
> norman-k-williams at uiowa.edu> wrote:
>
> Currently, itkImageIOBase contains what many people would regard as a
> terrible offense against the commandments of Generic Programming:
>
> bool ImageIOBase::SetPixelTypeInfo(const std::type_info &ptype);
>
> This function is called with the type_info for the pixel type from
> itk::ImageFileWriter.
>
> The problem is described here:
> http://public.kitware.com/Bug/view.php?id=11426
>
> where the complaint is that ImageIO can't write images with vector pixels
> longer than 7 elements.  The reason this is the case is that there's a long
> if/then/else chain that compares the ptype to the typeid of a set of
> explicitly named types, and no one, so far, had added vectors of length > 7
> elements.
>
> This would be a great place to do some Template MetaProgramming, but it's
> complicated.  Given the ImageType::PixelType one has to discriminate the
> PixelType (e.g. scalar,rgb,vector,covariantvector, complex), the number of
> components, and the component type.
>
> I'm feeling overwhelmed trying to sort this one out; I'm submitting a topic
> to gerrit that just punts -- i.e. I increase the possible vector size
> supported without re-organizing and template-izing things.
>
> Anyone want to give it a go?
>
> _______________________________________________
> Powered by www.kitware.com <http://www.kitware.com>
>
> Visit other Kitware open-source projects at
> http://www.kitware.com/opensource/opensource.html
>
> Kitware offers ITK Training Courses, for more information visit:
> http://kitware.com/products/protraining.html
>
> Please keep messages on-topic and check the ITK FAQ at:
> http://www.itk.org/Wiki/ITK_FAQ
>
> Follow this link to subscribe/unsubscribe:
> http://www.itk.org/mailman/listinfo/insight-developers
>
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/mailman/private/insight-developers/attachments/20101113/e3d5e6bf/attachment.htm>