Optimize quality computation

[zhiltsov-max]

Motivation and context

Most noticeable in tasks with large images (e.g. 2k x 4k pixels) and with ellipse annotations. The result is that for a sample task with 157 GT frames and with lots of small ellipse annotations, peak memory consumption falls from 20GB to under 2GB.

• Added lazy ellipse mask / RLE computation (OpenCV's ellipse() is quite slow with big masks)
• Removed simultaneous mask materialization for all masks on the image in segmentation matching
• Added batch size restriction in bulk saving of quality reports

How has this been tested?

Checklist

• I submit my changes into the develop branch
• I have created a changelog fragment
• I have updated the documentation accordingly
• I have added tests to cover my changes
• I have linked related issues (see GitHub docs)

License

• I submit my code changes under the same MIT License that covers the project.
  Feel free to contact the maintainers if that's a concern.

[sonarqubecloud]

Quality Gate passed

Issues
1 New issue
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 47.61905% with 11 lines in your changes missing coverage. Please review.

Project coverage is 73.40%. Comparing base (68ffe4a) to head (6a35a0f).
Report is 8 commits behind head on develop.

Additional details and impacted files

@@             Coverage Diff             @@
##           develop    #8990      +/-   ##
===========================================
- Coverage    73.83%   73.40%   -0.44%     
===========================================
  Files          417      415       -2     
  Lines        44602    44193     -409     
  Branches      4031     3859     -172     
===========================================
- Hits         32934    32439     -495     
- Misses       11668    11754      +86     

Components	Coverage Δ
cvat-ui	77.40% <ø> (-0.87%)	⬇️
cvat-server	70.07% <47.61%> (-0.01%)	⬇️

[Eldies]

LGTM