Pixelwise annotation tool


I want to do Pixelwise annotation for confocal images of cells.
Before starting new tools I wanted to check if there is an existing tool.

Can you please recommend me tools?

Hi and welcome!

I assume that your data is 3D since you mention confocal? For 2D you have more options, but what comes to my mind for 3D:

  • Labkit in Fiji . Skip the “training” and just use it for annotating.
  • If you are comfortable with python the same can be easily done with Napari, where you have e.g. additionally the flexibility to add own code callbacks to keyboard keys.

Both tools don’t provide any slice interpolation as far as I know.

  • There is Paintera as well. Don’t know it myself but I think rather powerful, steeper learning curve, interpolation options.
  • If it’s just a few cells you can also use the good old Segmentation Editor in Fiji. Easy to learn, not super fancy, interpolation options.

in general I’d say 3D annotation is labor intensive, because the tools are just not so very convenient yet.

PS: Consider whether you want to do semantic (binary) or instance (label id) annotation before starting.


Agreed, and if the original image is 3D, I would recommend asking about “voxel” annotations :slight_smile:

If you want 2D annotations, another option is QuPath which has both pixel classifiers (set a threshold, or train a machine learning classifier) and a nicely adjustable wand tool, in addition to other standard shapes (brushes).
One nice advantage is that the project structure is easy to maintain, and you can mass export resulting areas/masks if needed. Or simply export the coordinates of objects in an open source friendly Gson/Json format.


If you’re still looking for an annotation tool, Innotescus is a new browser-based annotation tool that supports robust instance and semantic segmentation tools on images and videos. We’re in beta mode so everything is free, and we’re planning on keeping it free for academic use - you can request a beta account here if it looks like it’ll work for you :slight_smile: