I work with CT images from rocks. I perform a segmentation in a CT volume then I have a 3d Numpy array with 3 labels (1,2,3) this numpy array is np.unit8. Then I transform this numpy array to ndrr format with this code?:
filename = 'post_process_fabrics_seg.nrrd'
# Write to a NRRD file
nrrd.write(filename, arr_seg)
I read the file as a volume using a 3D slicer. But I need to see the 3D volume label by label I only can visualize with 2d View. I apply some of the presets it seems that it only works for medical data.
Here is what I see:
You can load the volume as segmentation if you choose “Description” → “Segmentation” in the “Add data” window. You can also rename the file to post_process_fabrics.seg.nrrd to load it as segmentation by default.
For segments that have a small number of voxels it is ok we can see it as you see above.
For segments that have more voxels, it has a lot of voxels in the wall of the image and it does not permit to see inside the image. I try opacity but is there another way to do this? like to empathize more with the voxels that are the interior of the image?
You can try volume rendering for visualization. For that, you can either load the image as labelmap volume (as you did originally) or export the segmentation to binary labelmap. You can then drag-and-drop the labelmap from the tree in Data module to a 3D view to show volume rendering. It is easier to dynamically crop a volume rendering and opacity adjusts the volumetric density (and not the opacity of the boundary surface as for segmentation visualization).