smpy.filters.starlet.starlet_transform_2d

smpy.filters.starlet.starlet_transform_2d(data, nscales)

Compute the isotropic undecimated wavelet transform of an image.

Perform the starlet transform decomposition using the à trous algorithm to generate wavelet coefficients at multiple scales.

Parameters:

• data (numpy.ndarray) – Input 2D image.

• nscales (int) – Number of wavelet scales to compute.

Returns:

wavelet_bands – 3D array containing wavelet coefficients with shape (nscales, ny, nx) where the first nscales-1 bands are wavelet coefficients and the last band is the coarse approximation.

Return type:

numpy.ndarray

Notes

The starlet transform is an isotropic undecimated wavelet transform that preserves translation invariance. Each wavelet band represents the detail information between consecutive smoothing scales.