Annotation Web is a web-based annnotation system made primarily for easy annotation of image sequences such as ultrasound and camera recordings. It uses mainly django/python for the backend and javascript/jQuery and HTML canvas for the interactive annotation frontend.
Annotation Web is developed by SINTEF Medical Technology and Norwegian University of Science and Technology (NTNU), and is released under a permissive MIT license.
You are more than welcome to contribute to this project, and feel free to ask questions.
Presentation of Annotation Web on IEEE IUS 2021.
Please cite this article if you use annotation web in your work.
Annotation Web-An open-source web-based annotation tool for ultrasound images
Erik Smistad, Andreas Østvik, Lasse Løvstakken
2021 IEEE International Ultrasonics Symposium (IUS)
- Pure web-based system. Annotaters only need a web browser, they don't have to install anything, and everything is stored on the server.
- Fast and interactive annotation of temporal data/video using javascript and HTML 5 canvas.
- Secure login with two-factor authentication.
- Multiple annotation tasks are implemented, such image classification, segmentation using splines, landmark and bounding box.
For more information, see the following wiki pages:
- System overview - Overview of the system design, goal, code structure, etc.
- Development setup - Running annotation web locally on your machine
- Server setup - Running annotation web on a server
- Importing data
- User management - Create users, admins, change password, add 2FA for users ++
- Setting up an annotation task
- Exporting annotated data