WienerDeconvolution

Purpose

WienerDeconvolution runs SAIRPICO Wiener deconvolution in 2D, 2D Slice, or 3D mode. Use it as a regularized linear deconvolution baseline when a sigma-based 2D approximation or a measured/synthetic 3D PSF is available.

Inputs

• input_image: intensity image in png, tif, or tiff format.

• deconvolution_type: one of 2D, 2D Slice, or 3D. Default 2D.

• sigma: Gaussian PSF width for non-3D modes. Default 1.5.

• psf_image: existing volumetric PSF image required for 3D mode.

• regularization_lambda: regularization value passed as CLI -lambda. Default 0.01.

• padding: whether to process border pixels with padding. Default False.

Outputs

• output_image: deconvolved image. The default template is {input_image.stem}_deconvolved{ext}.

Assumptions

• The chosen mode matches the input image dimensionality.

• 3D mode uses psf_image; non-3D modes use sigma.

• The regularization scale is appropriate for the image intensity range.

Dependencies and Core Libraries

• BioImageFlow core processing classes.

• SAIRPICO simglib environment with bioimageit::simglib==0.1.2.

• External commands: simgwiener2d, simgwiener2dslice, and simgwiener3d.

Minimal Example

from bioimageflow_core import Arguments
from bioimageflow_sairpico_tools import WienerDeconvolution

result = WienerDeconvolution().process_row(
    Arguments(
        input_image="input.tif",
        deconvolution_type="2D",
        sigma=1.5,
        psf_image=None,
        regularization_lambda=0.02,
        padding=False,
        output_image="wiener.tif",
    )
)

Expected Results

The tool writes the deconvolved image and returns its path. For the example, the command starts with simgwiener2d and includes -sigma 1.5 and -lambda 0.02.

Failure Modes

• Missing SAIRPICO command or incompatible runtime platform.

• Missing PSF file for 3D mode.

• Non-zero subprocess exit due to unsupported image data or parameters.

• Output path cannot be created or written.