deepmr.prox.llr_denoise

deepmr.prox.llr_denoise(input, ndim, ths, W, S=None, rand_shift=True, axis=None, device=None)

Apply Local Low Rank denoising.

The solution is available in closed-form, thus the denoiser is cheap to compute.

deepmr.prox.ndim

Number of spatial dimensions.

Type:

int,

deepmr.prox.W

Patch size (assume isotropic).

Type:

int

deepmr.prox.ths

Denoise threshold. The default is 0.1.

Type:

float, optional

deepmr.prox.S

Patch stride (assume isotropic). If not provided, use non-overlapping patches.

Type:

int, optional

deepmr.prox.rand_shift

If True, randomly shift across spatial dimensions before denoising.

Type:

bool, optional

deepmr.prox.axis

Axis assumed as coefficient axis (e.g., coils or contrasts). If not provided, use first axis to the left of spatial dimensions.

Type:

bool, optional

deepmr.prox.device

Device on which the wavelet transform is computed. The default is None.

Type:

str, optional

Returns:

output – Denoised image of shape (…, n_ndim, …, n_0).

Return type:

np.ndarray | torch.Tensor