Add functionality to save stitched image as h5 or zarr

[IgorTatarnikov]

Description

What is this PR

• Bug fix
• Addition of a new feature
• Other

Why is this PR needed?
There is no way to save the fused image in the preview as one 3D image.

What does this PR do?
Adds functionality to save the fused image as either a BigDataViewer compatible h5 file or as an OME-Zarr.

How has this PR been tested?

Tests have been added to cover the new functionality.

NOTE:
To test the functionality locally you must install Fiji and also add BigStitcher to the update sites. This can be done using the GUI see here or via the command line after installing Fiji.

./ImageJ-linux64 --update add-update-sites "BigStitcherUpdate" "https://sites.imagej.net/BigStitcher/"
./ImageJ-linux64 --update update

Replace ImageJ-linux64 with the path to the local ImageJ executable.

Three different datasets are available for testing:

1. Tiny (~28 MB): this is mostly used for automated testing and doesn't work for visual inspection. It can be downloaded here.
2. Small (~1.8 GB): this dataset is more useful for visual inspection. It can be downloaded here.
3. A full size dataset is available (~130 GB): this dataset works best for testing overall performance, especially when reading or writing data. It's available by request.

Is this a breaking change?

No.

Does this PR require an update to the documentation?

No.

Checklist:

• The code has been tested locally
• Tests have been added to cover all new functionality (unit & integration)
• The code has been formatted with pre-commit

[codecov]

Codecov Report

Attention: Patch coverage is 91.84549% with 19 lines in your changes missing coverage. Please review.

Project coverage is 95.20%. Comparing base (58db8aa) to head (5e55376).

Files with missing lines	Patch %	Lines
brainglobe_stitch/stitching_widget.py	79.51%	17 Missing ⚠️
brainglobe_stitch/image_mosaic.py	98.66%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #14      +/-   ##
==========================================
- Coverage   96.89%   95.20%   -1.69%     
==========================================
  Files           6        6              
  Lines         419      626     +207     
==========================================
+ Hits          406      596     +190     
- Misses         13       30      +17     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[JoeZiminski]

Hey Igor this looks / works great! It ran nicely on the big_test_data and was easy to use. The code was clear to read, I'm not so familiar with zarr / h5 so could not pass an expert eye over everything but looks good to me. I've got some questions / suggestions on the code and below is some feedback from using the GUI:

• Once 'Fuse' is run, does it make sense to auto-load it into napari? Or maybe a popup like 'fused file saved to XXXX'. I wasn't sure what had happened once 'fused' was pressed. However, later learnt to keep an eye on the terminal when using napari. But maybe new users are not familiar with this and using it as a pure GUI.
• When you click the 'Browse' button, even if you don't select anything it overwrites anything you have in the lineedit. Sometimes I miss-clicked on it and it removed what I had in the lineedit when I didn't want it to. Maybe it is worth checking if the file browser dialog returns "" or whatever it's null return is, and in this case do not update the lineedit.
• I ran this on the very tiny dataset and got an error, might be worth looking into (but maybe just a small dataset issue)
• If I tried to save the fuse output to a folder that already exists, I get a relative uninformative error path '0' contains an array.

[JoeZiminski]

I am not very familiar with this setup but this looks good to me. Is it worth adding a test that all elements are read and assigned correctly? maybe this is overkill

[IgorTatarnikov]

Shouldn't be too hard, I think I have all of the expected values stored.