Skip to content

Installation

Technical Issues

The renaming process is ongoing, and you might encounter technical issues The new conda-forge package is not yet released, until then the in-app update does not work. Issues with the windows installer have been resolved.
If you encounter issues, please use PlantSeg <= 2.0.0rc14 for now, thank you!

This is the installation guide for the latest PanSeg. Please check the installation guide for PanSeg v1 at PanSeg Legacy Installation.

Download

Download the installer from heibox and choose according to your platform.

Download

Installation

The installer comes complete with its own python installation. During the installation further dependencies need to be downloaded.

First Start

The first start can be quite slow, as the machine learning models need to be downloaded.

Download the installer, then open a terminal and run the installer:

bash PanSeg*.sh

On linux, you can start PanSeg through your start menu.

To start PanSeg in the terminal, navigate to your installation directory (default ~/panseg/) and run bin/panseg --napari.

You might want to add a link to this file to some directory on your $PATH.

Alternatively, you can activate the panseg conda environment:
(Replace [INSTALLATION DIR] with the absolute(!) path to your installation)

eval "$("[INSTALLATION DIR]/bin/conda" shell.bash activate "[INSTALLATION DIR]")
panseg --help

Download the installer and execute it. As the installer is not signed you might need to confirm the download and that you want to run the installer.

Choose a installation path without spaces in it, as those can cause issues with conda packages.

Start PlanSeg through the Windows start menu.

If you want to install PanSeg without the installer, you need to have conda. (We recommend Microforge to get conda, see installing mamba)

conda create --name panseg panseg
conda activate panseg
panseg --napari

To get the latest pre-release features, install PanSeg from git. You need to have conda and git installed. (We recommend Microforge to get conda, see installing mamba)

git clone https://github.com/kreshuklab/panseg.git
cd panseg
conda env create -f environment.yaml
conda activate panseg
panseg --napari

For development, we recommend using the environment-dev.yaml instead! Also check the contributing section

Updating

Info

Due to an external change, this only works from version 2.0.0rc5 onward.
If you are running an older version, please uninstall and reinstall PanSeg.

If you have installed PanSeg via the installer or from Conda-forge, you can update to a new version right in the GUI! Go to the Plugins menu on top, then click Update Panseg!

If you have cloned the git repository, you need to update your local repo:

cd panseg
git pull
conda env update -f environment.yaml # environment-dev.yaml for development

Optional dependencies

Certain compressed TIFF files (e.g., Zlib, ZSTD, LZMA formats) require additional codecs to be processed correctly by PanSeg. To handle such files, install the imagecodecs package:

conda activate panseg
pip install imagecodecs

If you plan to use SimpleITK-based watershed segmentation, you will need to install SimpleITK as an additional dependency:

conda activate panseg
pip install SimpleITK

Installing PanSeg v1

Please check the installation guide for PanSeg v1 at PanSeg Legacy Installation.