Improve Visualisation Capability in Scicat

[LAShemilt]

Visualisation Capability in SciCat

Summary

Currently visualisation can only be achieved through thumbnails and attachments. This sets limitations on the size and type of data that can be visualised. Furthermore, these details have to be manually added by users, they are not typically automatically ingested.

We propose some changes to help SciCat move towards supporting better visualisations.

User requirements

1. Adding videos , and images stacks as attachements
2. Adding interactive plots via plotly of both metadata and/or data
3. Using WebGL visualisation to provide interactive 3D rendering of data
4. Supporting annotation pipelines on visualisation.

Changes to SciCat Frontend

For 1 and 2, this could be supported through microfrontends that could be added as plugins. A repository for useful plugins could be set up in the SciCat project repo. The underlying data could be stored in the mongodb compressed or as URLS on a separate system that are drawn into SciCat.

For 3,4 would have to be supported by a separate microservice with a link out from SciCat using technology like neuroglancer. This would be data and process specific. There would be implications on how the underlying data is stored.

[Ingvord]

Hi, I have created a dedicated thread to discuss technicalities related to plugins system here

[Ingvord]

I took the liberty to draw some mindmap related to the visualization in SciCat.

Please let me know if that makes any sense or it is complete rubbish