[Insight-users] Re: CVS checkout

Jessica de Ryk jessica-deryk at uiowa . edu
Mon, 15 Dec 2003 12:15:18 -0600


I tried getting the filter from CVS. As I am a user I used this command to
get the checkout.
cvs -d:pserver:anoncvs at www . itk . org/cvsroot/Insight co Insight

Unfortunately the version I got doen't contain the
itkVectorResampleImageFilter or the related example,
ResampleImageFilter6.cxx

Do I need a different CVS command to get the newest update?
Thanks again
Jessica



----- Original Message ----- 
From: "Luis Ibanez" <luis . ibanez at kitware . com>
To: <jessica-deryk at uiowa . edu>
Cc: <insight-users at itk . org>
Sent: Friday, December 12, 2003 2:09 PM
Subject: Re: [Insight-users] Re: Images Posted


> Hi Jessica,
>
> ITK was missing a filter for resampling RGB images,
> which is one of the steps required for your program.
>
> We just added a new filter:
>
>          itkVectorResampleImageFilter
>
> Please update your CVS checkout in order to get this
> filter. You will find and example on how to use this
> filter in the file:
>
>
>    Insight/Examples/Filtering/ResampleImageFilter6.cxx
>
>
> Please use this filter intead of the plain itkResampleImageFilter
> in order to resample your RGB images.
>
> Note that you don't need 3 dimensions. In ITK the notion
> of color channels is represented using the RGBPixel class.
> You simply need a 2D image of RGBpixels, like:
>
>
>     typedef    itk::RGBPixel< unsigned char >   PixelType;
>     typedef    itk::Image< PixelType, 2 >       ImageType;
>
>
> Please let us know if you find any problems,
>
>
> Thanks
>
>
>    Luis
>
>
> ------------------------------------
> jessica-deryk at uiowa . edu wrote:
>
> > I am composing a mosaic of many small 2D RGB image tiles to make a
single
> > image of a complete slice of lung. The images are stereomicroscopic not
> > confocal (but one day the system will hopefully use confocal). A 3D
> > reconstruction of the tissue will be generated from the many 2D slices
and the
> > resulting pathology dataset registered with corrosponding CT images.
> >
> > I do just need to resample the RGB image. I thought I might have needed
three
> > dimensions (for the three channels) I guess i never tried resampling the
RGB
> > images with the 2D resample.
> >
> > If i use the ImageReadWrite example, is there a way inwhich I can
specify my
> > pixel spacing or should I modify the text file once it is generated?
> >
> > Thanks again
> > Jessica
> >
> > Quoting Luis Ibanez <luis . ibanez at kitware . com>:
> >
> >
> >>Hi Jessica,
> >>
> >>Thanks for posting your images in MyPACS.
> >>It was pretty easy to look at them.
> >>
> >>I'm confused now if you are using 3D images
> >>like in confocal microscopy or you are just
> >>composing the mosaic of a large image by
> >>stitching the tiles in rows and columns
> >>in a 2D grid. Could you please clarify this ?
> >>
> >> From your last email it looks like you just
> >>need to resample the RGB 2D image...
> >>is that right ?
> >>
> >>----
> >>
> >>If you want to convert the images from one format
> >>to another you can simply use the ITK example
> >>
> >>    Insight/Examples/IO/ImageReadWrite.cxx
> >>
> >>you can for example read a PNG image and write it
> >>as Analyze or MetaImage format.
> >>
> >>I would suggest you to use MetaImage since the
> >>header is in ascii text and it is easy to read.
> >>
> >>Simply execute the program as:
> >>
> >>    ImageReadWrite  input.png   output.mhd
> >>
> >>That will convert your input PNG image into
> >>a MetaImage format. You can also read JPEG
> >>and TIFF images (if you are using the CVS
> >>version of ITK).
> >>
> >>---
> >>
> >>Regards,
> >>
> >>
> >>   Luis
> >>
> >>
> >>----------------------
> >>Jessica de Ryk wrote:
> >>
> >>
> >>>I have built the adjusted code but am having run time problems when I
> >>>input my images. I get an abnormal program termination (exception
> >>>breakpoint) I have posted an example of two images on the MyPACS.net
> >>>(the case number is 253148)
> >>>Maybe I am aproaching final output of producing the cropped tile
> >>>in wrong way? I have many images to tile together to construct a single
> >>>color slice (the 3D was for the color as opposed to grayscale) of lung
> >>>pathology. The overlap is 20% of the image FOV + /-some error caused by
> >>>inaccuracy in the aquisition system. I want to register two images and
> >>>then join them to produce one aligned image. Then register the aligned
> >>>image with the next tile, etc first building up image rows then
> >>>registering the individual rows. I thought this would be very straight
> >>>forward to accomplish. As I am much more familiar with C, I thought I
> >>>could program the organization of the images with respect to eachother,
> >>>and the overall tiling of the images and call my itk based registration
> >>>program to produced the correctly cropped image tiles.
> >>>What program would you recomend to convert my images?? Which format is
> >>>better, Analyze or MetaImage
> >>>Thanks for your ongoing advice
> >>>Jessica
> >>>
> >>>
> >>>
> >>>
> >>
> >>
> >>
> >
> >
> >
> >
> > _______________________________________________
> > Insight-users mailing list
> > Insight-users at itk . org
> > http://www . itk . org/mailman/listinfo/insight-users
> >
>
>
>