Fix distance_transform indexing for non-square inputs

[omkar-334]

What does this PR do?

Fixes #3300

Before submitting

• Was this discussed/agreed via a Github issue? (no need for typos and docs improvements)
• Did you read the contributor guideline, Pull Request section?
• [] Did you make sure to update the docs?
• Did you write any new necessary tests?

Fixes an indexing bug in the PyTorch implementation of distance_transform that affected non-square 2D inputs, and caused hausdorff_distance to fail for shapes where height is greater than width.

Tests

Added 2 tests for non-square inputs:

• distance_transform on an 11x10 tensor, compared against SciPy for:
  • euclidean
  • chessboard
  • taxicab
• hausdorff_distance on shape (1, 1, 11, 10), compared against MONAI

---

📚 Documentation preview 📚: https://torchmetrics--3338.org.readthedocs.build/en/3338/

[omkar-334]

Before:

After:

[omkar-334]

New tests passing:

[codecov]

Codecov Report

❌ Patch coverage is 0% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 32%. Comparing base (c781c99) to head (7853799).
⚠️ Report is 1 commits behind head on master.

❗ There is a different number of reports uploaded between BASE (c781c99) and HEAD (7853799). Click for more details.

HEAD has 2 uploads less than BASE

Flag	BASE (c781c99)	HEAD (7853799)
gpu	1	0
unittest	1	0

Additional details and impacted files

@@           Coverage Diff            @@
##           master   #3338     +/-   ##
========================================
- Coverage      37%     32%     -5%     
========================================
  Files         364     349     -15     
  Lines       20096   19901    -195     
========================================
- Hits         7521    6377   -1144     
- Misses      12575   13524    +949     

🚀 New features to boost your workflow:

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

[Copilot]

Pull request overview

Fixes an indexing bug in the PyTorch distance_transform implementation that breaks for non-square 2D inputs (notably when height > width), which in turn caused hausdorff_distance to error for certain shapes (issue #3300).

Changes:

• Corrected flatten indexing in distance_transform for non-square inputs.
• Added a new unit test covering non-square distance_transform outputs vs SciPy across distance metrics.
• Added a regression test for hausdorff_distance on (1, 1, 11, 10) vs MONAI.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

File	Description
src/torchmetrics/functional/segmentation/utils.py	Fixes distance_transform flatten indexing by using width for row-major indexing.
tests/unittests/segmentation/test_utils.py	Adds SciPy-based regression test for non-square distance_transform inputs.
tests/unittests/segmentation/test_hausdorff_distance.py	Adds MONAI-based regression test for non-square hausdorff_distance inputs.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

You can also share your feedback on Copilot code review. Take the survey.

[omkar-334]

@justusschock all tests are passing.... can you review this?