[Copilot]

cv2.aruco.ArucoDetector.detectMarkers() returns marker IDs as either (N,1) elements or scalar-like (N) elements depending on OpenCV/platform; detect_markers() assumed array indexing (marker_id[0]) and crashed on macOS with IndexError. This PR normalizes IDs at use-site and hardens CI output/reporting so matrix test failures are immediately diagnosable.

• Marker ID Compatibility in detect_markers

  • Added _normalize_marker_id() to coerce detector outputs (7, np.array(7), np.array([7]), np.array([7])[0]-like) into a single int.
  • Replaced direct marker_id[0] usage with the normalized value for both:
    • "marker id"
    • "marker name" ({family}_{id})

• Focused Regression Coverage

  • Added a dedicated test module for marker ID normalization behavior.
  • Covers scalar and one-element-array representations and asserts the emitted detection fields remain correct (marker id, marker name).

• CI Diagnostics in .github/workflows/main.yml

  • Updated pytest invocation to include verbose test names and long tracebacks (-vv --tb=long).
  • Added JUnit XML output for every matrix run.
  • Added upload-artifact with if: always() so reports are preserved even on pytest failure.
  • Artifact names include OS/Python dimensions to avoid collisions.
  • Kept Windows MSMF override (OPENCV_VIDEOIO_PRIORITY_MSMF: "0") and used OS-appropriate shell commands for report directory setup.

def _normalize_marker_id(marker_id: int | np.ndarray) -> int:
    return int(np.asarray(marker_id).reshape(-1)[0])

# in detect loop
marker_id_int = _normalize_marker_id(marker_id)
record = {
    "marker id": marker_id_int,
    "marker name": f"{fam_name}_{marker_id_int}",
}

[qian-chu]

LGTM!

Disclosure of Delegation to Generative AI

The authors declare the use of generative AI in the research and writing process. According to the GAIDeT taxonomy (2025), the following tasks were delegated to GAI tools under full human supervision:

• Code generation
• Code optimization
• Validation

The GAI tool used was: GPT-5.4.
Responsibility for the final manuscript lies entirely with the authors.
GAI tools are not listed as authors and do not bear responsibility for the final outcomes.
Declaration submitted by: Qian Chu