Seg2linkUnet2d is a package for training a 2D U-Net to predict cell/non-cell regions in 3D bio-medical images.

Seg2linkUnet2d and Seg2Link can be used together to perform semi-automatic 3D cell segmentation. Read the documentation to learn how to do it.

Install

• Install Anaconda or Miniconda
• Create a new conda environment and activate it by:

$ conda create -n unet2
$ conda activate unet2

• Install PyTorch
• Install seg2link_unet2d:

$ pip install seg2link-unet2d[local]

Use the notebook for training/segmentation

• Activate the created environment by:

$ conda activate unet2

• Start the jupyter notebook

$ jupyter notebook

• Open the notebook under the /Examples folder and run the codes according to the instructions.