xpunwrap.algorithms.algo_skimage_unwrap

xpunwrap.algorithms.algo_skimage_unwrap(phase_wrapped: ndarray, restore_plane: bool = False, **kwargs: Any) → ndarray[source]

2D phase unwrapping using scikit-image’s CPU implementation.

Parameters:

  • phase_wrapped (xp.ndarray) – Wrapped phase, shape (H, W) or (N, H, W), values in [-pi, pi).

  • restore_plane (bool, optional) – Ignored for this algorithm; scikit-image already preserves the global ramp (Poisson null-space is resolved internally). Kept for interface compatibility. Default False.

  • kwargs – Passed through to skimage.restoration.unwrap_phase (e.g. spacing, wrap_around, seed).

Returns:

Unwrapped phase, same shape and backend as the input.

Return type:

xp.ndarray

Notes

  • Requires scikit-image to be installed in the active environment.

  • Computation is performed on CPU; CuPy inputs are transferred to NumPy and converted back to the active backend on return.

  • Unlike the FFT-based solvers in this package, scikit-image’s unwrap_phase uses a path-following (quality-guided) algorithm that does not assume periodic boundary conditions and does not use FFTs. Pass wrap_around=(True, True) via kwargs if your data has periodic boundaries.