Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Extracting the current VoxelGrid for an additional processing step #157

Open
rcffc opened this issue Apr 16, 2021 · 1 comment
Open

Extracting the current VoxelGrid for an additional processing step #157

rcffc opened this issue Apr 16, 2021 · 1 comment

Comments

@rcffc
Copy link

rcffc commented Apr 16, 2021

Hi everyone.
My goal is to add an extension that performs semantic instance segmentation on the Voxel Grid.

Therefore, I need to

  1. extract the current reconstructed scene as voxelgrid
  2. map the segmentated voxels back into the reconstructed scene

Is there a method to access the global voxel block?

Any help is appreciated.

@Dennis199607
Copy link

Hey, I think I can help you with that since I am also working with Infinitam for my master thesis.

You need to take a look into the ITMRepresentationAccess.h file. In the project you find it in ITMLib/Objects/Scene/ITMRepresentationAccess.h. There are several methods to directly access voxels in your scene.

I think for accessing a voxel I used 'readVoxel(localVBA, hashTable, globalBlockLocation, vmIndex)'

localVBA is the pointer to the local Voxelblock Array (you can get it in ITMScene in localVBA.GetVoxelBlocks()).
hashTable is the pointer to the Hash Table (you can get it also in ITMScene in index.getEntries())
globalBlockLocation is just a Vector3i, the global position of the voxel
vmIndex is an integer, which tells you after the call whether it found an allocated entry for that voxel (1) or not (0)

If you need more help we can also get in contact.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants