xfeatures2d: fix matchLOGOS crash with fewer than N neighbors

[JaimeFine]

Fixes #4174

• Fix out-of-bounds access in Point::nearestNeighboursNaive() when fewer than N neighbors are available.
• Resize nnVector to the actual neighbor count to avoid invalid trailing neighbor entries.
• Add a regression test for the 3-keypoint same-label matchLOGOS reproducer.

Tested:

• Built opencv_test_xfeatures2d on Windows with OpenCV 5.x + opencv_contrib 5.x
• Ran LOGOS-related tests with OPENCV_TEST_DATA_PATH set to C:\dev\opencv_extra\testdata

Pull Request Readiness Checklist

See details at https://github.com/opencv/opencv/wiki/How_to_contribute#making-a-good-pull-request

• I agree to contribute to the project under Apache 2 License.
• To the best of my knowledge, the proposed patch is not based on a code under GPL or another license that is incompatible with OpenCV
• The PR is proposed to the proper branch
• There is a reference to the original bug report and related work
• There is accuracy test, performance test and test data in opencv_extra repository, if applicable
  Patch to opencv_extra has the same branch name.
• The feature is well documented and sample code can be built with the project CMake

NOTE: No new test data or performance test is needed for this bug fix; the regression test uses synthetic keypoints.

[JaimeFine]

The failing GitHub Actions run appears to be unrelated to this patch.

The workflow parser fails before running tests because
.github/workflows/PR-5.x.yaml references a reusable workflow with @geometry_module,
which does not appear to be a valid branch/tag/commit in opencv/ci-gha-workflow?

Local validation for this patch completed successfully.

[JaimeFine]

Hi OpenCV maintainers, gentle follow-up on this small xfeatures2d bug fix.

This PR fixes #4174 by avoiding an out-of-bounds access in matchLOGOS when fewer than N candidate neighbors are available, and adds a synthetic regression test for the small-neighbor case.

I noticed the GitHub Actions workflow did not reach the test stage because of what looks like an unrelated reusable-workflow reference issue. I validated locally on Windows with OpenCV 5.x + opencv_contrib 5.x by building opencv_test_xfeatures2d and running the LOGOS-related tests.

Please let me know if there is anything I should adjust to make the PR easier to review. Also happy to retarget or rebase if 5.x is not the preferred branch for this fix.

Thanks!