feat(compact_mask): add resize() method and benchmark stage

[Borda]

This pull request adds a new resize method to the CompactMask class, enabling efficient resizing of masks to arbitrary resolutions with nearest-neighbor interpolation. The benchmark and summary reporting code is updated to include timing and correctness checks for the new operation, and a comprehensive suite of tests is provided to ensure correctness and robustness of the resize logic.

CompactMask resize implementation and tests:

• Added a resize method to CompactMask, which rescales masks, offsets, and crop shapes to a new image size using nearest-neighbor interpolation and updates the internal RLE encoding. The method handles identity, empty, and invalid-resize cases efficiently and raises an error for non-positive dimensions.
• Added a new test class TestCompactMaskResize with extensive parameterized and property-based tests covering proportional scaling, identity, empty masks, invalid shapes, non-square scaling, extreme downscaling, and pixel-level parity with OpenCV's resizing.

Benchmark and reporting updates:

• Implemented a new stage_resize timing and correctness check in the benchmark, using the new resize method and a dense mask baseline, with 1-pixel tolerance for correctness. Integrated these metrics into scenario results and reporting. [1] [2] [3] [4] [5]
• Updated the summary table, CSV export, and reporting logic to include the new "resize" operation, its timing, speedup, and correctness status. [1] [2] [3] [4] [5] [6]

Internal utilities:

• Added _resize_dense_to_shape as a utility for resizing dense masks using numpy fancy-indexing, to serve as a baseline for correctness and timing comparisons.

These changes ensure that CompactMask supports robust and efficient resizing, with full test coverage and integration into benchmarking and reporting workflows.

[codecov]

Codecov Report

❌ Patch coverage is 93.70079% with 8 lines in your changes missing coverage. Please review.
✅ Project coverage is 78%. Comparing base (87e8153) to head (0fbc48c).

❌ Your project check has failed because the head coverage (78%) is below the target coverage (95%). You can increase the head coverage or adjust the target coverage.

Additional details and impacted files

@@           Coverage Diff            @@
##           develop   #2227    +/-   ##
========================================
  Coverage       77%     78%            
========================================
  Files           66      66            
  Lines         8211    8338   +127     
========================================
+ Hits          6350    6469   +119     
- Misses        1861    1869     +8     

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.