search

SCIInstitute / Seg3D

Seg3D is a free volume segmentation and processing tool developed by the NIH Center for Integrative Biomedical Computing at the University of Utah Scientific Computing and Imaging (SCI) Institute.
109 stars 47 forks source link

Can't render images larger than 2^14 by 2^14 pixels #249

Open blakezim opened 6 years ago

blakezim commented 6 years ago

The program can render images that are at or below 2^14 pixels in any dimension. The images I was using was 2D, but I would also like to exceed this limit in 3 dimensions, but I don't have an example case for that. I have included some screen shots. I was using Seg3D 2.3.2, but I have also tried this with Seg3D 2.4.2 and I experienced the same issue. My OS is opensuse 42.2. If a test image is need, please ask and I can supply one.

not_working_crop working_crop

jessdtate commented 6 years ago

This could be a hard ware limitation with opengl texture limitations.

blakezim commented 6 years ago

Yea, that is what we realized. There is a way with opengl to tile the textures so that it doesn't exceed the limit for any one tile (at least that was what I understood). But I understand that this could be quite the restructure.

jessdtate commented 6 years ago

using the large volume tools should downsample the image to something it can view. did that work?

blakezim commented 6 years ago

I tried that before and it didn't work either, which I would assume (my guess) means it is still trying to use one texture that is overlaid on all of the tiles that Seg3D created during the large volume processing. When I hover over the image with the mouse, it know what value is there and it displays it in the bottom right corner, but it just doesn't show up.

allywarner commented 5 years ago

Hi Blake, could you include an example image here please?

blakezim commented 5 years ago

The path to some testing images is the following:

/home/sci/blakez/Issues_Seg3D/Issue_249/testing_images/

Best,

Blake Zimmerman Graduate Research Assistant Scientific Computing and Imaging Institute (SCI) Utah Center for Advanced Imaging Research (UCAIR) Department of Bioengineering, University of Utah

On Wed, Mar 27, 2019 at 2:36 PM allywarner notifications@github.com wrote:

Hi Blake, could you include an example image here please?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/SCIInstitute/Seg3D/issues/249#issuecomment-477337245, or mute the thread https://github.com/notifications/unsubscribe-auth/AYahZw13nPgmGonNtSVrIdqtODNY2jIgks5va9Y2gaJpZM4R79eu .

allywarner commented 5 years ago

/home/sci/blakez/Issues_Seg3D/Issue_249/testing_images