AtlasSpotDetection

AtlasSpotDetection wraps the external Atlas spot detection CLI. It detects sparse spots in 2D intensity TIFF images and writes a binary detection mask.

Use this tool when a workflow specifically needs the Atlas CLI behavior. For lightweight deterministic puncta detection without the external Atlas binary, use DetectSpots.

Inputs are input_image, optional gaussian_std, optional p_value, optional area_lim, and verbose. Output is output_image, a binary TIFF detection mask. The tool requires an ExecutionContext with row_dir because the Atlas CLI writes implicit files in the process working directory.

Core dependencies are BioImageFlow core APIs and the external bioimageit::atlas conda package. The wrapper also uses a packaged blobs.txt Atlas reference, with a generated shared fallback in the workflow work directory when packaged reference data is unavailable.

from bioimageflow_spot_tools import AtlasSpotDetection

spots = AtlasSpotDetection()(
    input_image=image["output_image"],
    p_value=0.05,
    gaussian_std=2,
    name="atlas_spots",
)

Expected result: output_image points to a binary mask where non-zero pixels represent detected spots.

Dependencies and Core Libraries

BioImageFlow core APIs, the external Atlas CLI, and the bioimageit::atlas conda package.

Assumptions

The input is a 2D TIFF intensity image and execution happens inside a BioImageFlow row context with a writable row work directory.

Failure Modes

Missing Atlas or blobsref binaries, missing row context, unsupported inputs, inability to write the shared reference, or non-zero CLI exit status stop execution.