Make a napari frontend prototype

[niksirbi]

napari has multiple built-in layer types which could be useful for movement.

• The Image layer could be used to display video frames
• The Points layer can be used to represent keypoints, see this example
• The Tracks layer can be used to visualise keypoint trajectories across time
• The Vectors layer can be used to show stuff like head orientation
• The Shapes layer can be used to create and show ROIs

Napari widgets can be created easily.

I will first make a prototype with the following features:

### Tasks
- [ ] https://github.com/neuroinformatics-unit/movement/issues/38
- [ ] https://github.com/neuroinformatics-unit/movement/issues/47
- [ ] https://github.com/neuroinformatics-unit/movement/issues/283

[niksirbi]

Work in progress in the napari branch.

[niksirbi]

Useful resources for designing napari plugins:

• napari plugin desing workshop slides
• list o popular napari plugins to try out

[niksirbi]

What should the prototype include

I have broken down the requirements for the prototype into simpler tasks and added them to this issue as a Tasklist.

[niksirbi]

The current work-in progress is on this branch, and is currently open as a draft pull request #112

[niksirbi]

While working on #112, I realised that it was all getting too complex for one pull request, so I'm shifting strategy.

We decided to proceed as following:

• Make a napari-dev branch to act as a temporary "main" branch for the development of the napari plugin + widget
• Break down the content of this PR into smaller, more manageable chunks and open them as individual PRs against napari-dev.

Once this process is well underway, and napari-dev catches up with what's here, I'll close PR #112 and delete the napari-pose-reader branch.

[sfmig]

I was just chatting to Sanna and we remembered she left valuable feedback re ROIs in the WAZP repo: I link it here in case it is useful

[niksirbi]

Yeah, and I implemented those suggestions in a PR I never merged.