3D Modelling & volume calculation

Hello, we apologise for the mistakes, English is not our native language.
We are novices on ImageJ.
We are working on a 3D modelling of igneous rocks from series of 2D images extracted from CT.
For that, we use Fiji.

Those images are in gray scale. We have to binarize our images (in order to isolate certain elements to calculate their volume).

  1. We’d like to know which filters fit the best to gray scale images and which function we could use to calculate volumes from binarized images?
  2. How can we know which is the space’s length between 2 images 2D ? And how can we change it ?

Many thanks for your help!

Hello @werocks and welcome to the ImageJ forum!

You may start by having a look at the thresholding methods. If the elements you are trying to isolate do not have distinct intensities compared to the background, you might need to use a more complex segmentation method.

You can do that for example using MorphoLibJ’s Particle Analysis 3D plugin.

In a 2D stack, that is called the Voxel depth, and it is set under Image > Properties…. You should be able to know its exact value based on your CT configuration.

1 Like

Thank you for your answer!

Yes, the difficulty is that we have to isolate elements which are closed in terms of grayscale values.
Plus, as our images are extracted from CT (computed tomography), we have a ring effect on our images.

Which methods do you think we can use?

We had another question:
Is it possible, once we have binarize our images (into black and white), to transform our selection (which is in black for example), into a 3D volume ? (in order to realize measurements on it after that)

You might want to give a try to the Trainable Weka Segmentation plugin in 3D. But be aware that working in 3D you might need quite a bit of RAM.

Regarding the ring effects, you might need to preprocess the image.

Yes, you can use the 3D viewer for that.