Hi Jess,

I am using File->Export Segmentation. Both 'to matlab' and 'to nrrd' have the same problem.

Here are some screenshots:
http://imgur.com/a/Y5knb

Image 1: `cap' visible outside image range.

Image 2: Exporting, see the cap is component 7.

Image 3: viewing in matlab, with hole clearly visible on the left.

As there does seem to be a component 7 in the matlab data, I've tried to see where it ends up, and turns out the problem is even worse than simple omission.
Component 7 gets translated to somewhere else in the volume and intersects with it, as shown here: http://imgur.com/a/tyCwd

(output of figure(1); clf; isosurface(scirunnrrd.data == 1,0.5); hold on; p=patch(isosurface(scirunnrrd.data ="=7,0.5));set(p,'FaceColor','r');" axis equal)

Thanks,

Sander

On Thu, Feb 4, 2016 at 4:57 PM, Jess Tate <jess@sci.utah.edu> wrote:

Hi Sander,

It’s a little tricky to understand what is going on, because when you export a segmentation, it should just save out the entire label mask.  Is it possible you are exporting the original image on accident?  You can try the 'Export Segmentation' option in the file menu, which has more controls than right clicking on the layer and exporting it that way.  If that isn’t working, a screen shot may help determine what is going on.

cheers,
Jess

> On Feb 4, 2016, at 5:05 AM, sander.land@gmail.com wrote:
>
> I have an image data set which is missing a small part due to the bounds of
> the imaging. I've filled this in using the paint tool to make a smooth shape.
> However, when exporting the data set it's cut off again at the original
> bounds. Is there a way to extend these bounds and export the entire
> segmentation?
>
> Thanks,
> Sander