Signatures module

Signatures module#

napari_shape_odyssey.signatures.heat_kernel_signature(surface: napari.types.SurfaceData, times: numpy.ndarray, order: int = 100, robust: bool = False, scaled: bool = True) → numpy.ndarray#

Compute the heat kernel signature of a surface.

Parameters

surface (napari.types.SurfaceData) – A napari surface tuple.
time (np.ndarray,) – Time values for which to calculate heat kernel values. Good values range over several orders of magnitude (1 - 1e7).
order (int, optional) – The order of shape spectrum to caluculate, by default 100
robust (bool, optional) – Use robust laplacian or not, by default False

Returns

np.ndarray – The heat kernel signature. Shape is (n_vertices, n_times).

Return type

np.ndarray

See also

[0] Bronstein & Kokkinos, “Scale-invariant heat kernel signatures for non-rigid shape recognition”, 2010, 10.1109/CVPR.2010.5539838

napari_shape_odyssey.signatures.wave_kernel_signature(surface: napari.types.SurfaceData, energies: numpy.ndarray, order: int = 100, sigma: float = 1, scaled: bool = False, robust: bool = False) → numpy.ndarray#

Compute the wave kernel signature of a surface.

Parameters

surface (napari.types.SurfaceData) – A napari surface tuple.
energies (np.ndarray) – Energy values for which to calculate wave kernel values.
order (int, optional) – The order of shape spectrum to caluculate, by default 100
sigma (float, optional) – The sigma to use for the wave kernel signature, by default 1
scaled (bool, optional) – Scale the wave kernel signature or not, by default False
robust (bool, optional) – Use robust laplacian or not, by default False

Returns

np.ndarray – The wave kernel signature. Shape is (n_vertices, n_energies).

Return type

np.ndarray