AIDA is an attempt to bring an open source web-based work-flow to image annotation. Currently, in the biomedical imaging space, image annotation is largely confined to single computer shrink-wrapped software with limited interactive capabilities and few, usually closed, data formats.
AIDA is a web interface that enables distributed teams of researchers to directly annotate images with easy to use on screen drawing tools. AIDA supports the creation of well defined annotation trials which include a series of high resolution images and a specific set of annotation tasks.
For documentation and further information see the Wiki.
Play with a live example here
The user interface is a VueJS Single Page Application, encapsulating and interacting with two other significant JavaScript libraries: OpenSeaDragon to manipulate the high-res images and PaperJS to provide the drawing functionality. VuetifyJS is used as a UI component library which implements the material design language.
AIDA reads and writes data in a simple JSON structure over a web API. Login and user accounts can enable specific configuration of the tool for different trials and tasks.
The software is published as Open Source under the permissive MIT license. The API will also be public.
The next stage of development will be to integrate intelligent tools that leverage the power of machine learning techniques. We hope to enhance the ability of the user to quickly and accurately mark up images through predictive assistance
You can use AIDA on your local machine. The only requirement NodeJS >v11.
A built and built version of the local application is included with the source code in the /dist directory. To begin using AIDA locally:
- Clone the repository
cdto/dist- Install dependencies via
npm install - Add the images you want to annotate to the
/dist/data/imagesdirectory. - Run the nodeJS local application via
node aidaLocal.js - Navigate to the localhost webserver specified in the console
- Annotations are read from and written to
/dist/data/annotations
Requirement NodeJS. Example work-flow:
- Clone the repository
- Install dependencies via
npm install - For development: start a hot-reloading dev server with
npm run serve - For deployment: bundle together with
npm run build
This interface was built by Alan Aberdeen. It is a project of Jens Rittscher, Nasullah Khalid Alham and Alan Aberdeen at the University of Oxford, specifically the Quantitative Biological Imaging Group.