https://github.com/qupath/qupath-extension-bioimageio
A QuPath extension for working with the Bioimage Model Zoo (https://bioimage.io)
https://github.com/qupath/qupath-extension-bioimageio
Last synced: about 1 year ago
JSON representation
A QuPath extension for working with the Bioimage Model Zoo (https://bioimage.io)
- Host: GitHub
- URL: https://github.com/qupath/qupath-extension-bioimageio
- Owner: qupath
- License: apache-2.0
- Created: 2022-11-29T08:19:20.000Z (over 3 years ago)
- Default Branch: main
- Last Pushed: 2025-02-12T16:41:04.000Z (over 1 year ago)
- Last Synced: 2025-03-23T06:41:29.413Z (about 1 year ago)
- Language: Java
- Size: 215 KB
- Stars: 4
- Watchers: 3
- Forks: 3
- Open Issues: 2
-
Metadata Files:
- Readme: README.md
- Changelog: CHANGELOG.md
- License: LICENSE
Awesome Lists containing this project
README
[](https://qupath.readthedocs.io/en/stable/docs/deep/bioimage.html)
[](https://forum.image.sc/tag/qupath)
[](https://github.com/qupath/qupath-extension-bioimageio/releases/latest)
[](https://github.com/qupath/qupath-extension-bioimageio/releases)
# QuPath Bioimage Model Zoo extension
This is a QuPath extension for working with the Bioimage Model Zoo (https://bioimage.io).
You can download the .jar file from [qupath-extension-bioimageio-0.1.0.jar](https://github.com/qupath/qupath-extension-bioimageio/releases/download/v0.1.0/qupath-extension-bioimageio-0.1.0.jar)
It's early and experimental - subject to change in later releases.
The main aim is to enable models kept in the Zoo to be imported into some QuPath-friendly form.
Currently, this extension adds a command to create pixel classifers from supported models:
* *Extensions → Bioimage Model Zoo → Create pixel classifier (Bioimage Model Zoo)*
Along the way, it adds some extra flexibility: choosing the resolution at which the model is applied, and adapting the input channels and output classifications.
Only a subset of models are currently supported, partly because
1. not all models are Java-friendly, and
2. some models require accessing all the pixels in the image... but QuPath is designed for huge images, and doesn't easily support these kinds of global calculations yet
As a result, applying the model in QuPath can sometimes give a different result to applying it in other software.
The goal is to improve this in future releases.