Skip to content

Annotation File to OME-TIFF

This example shows how to turn a volume annotation downloaded from WEBKNOSSOS into a OME-TIFF. When manually downloading a WEBKNOSSOS annotation through the UI you end up with a ZIP file containing the volume segmentation in the WKW-format (and potentially any skeleton annotation).

As an alternative to manually downloading annotation files, have a look at streaming the data directly from the remote serve, e.g., in this example.

from pathlib import Path

from tifffile import imwrite

import webknossos as wk

# Specify a bounding box for cutouts
# (topleft_x, topleft_y, topleft_z), (width, height, depth)
BOUNDING_BOX = wk.BoundingBox((0, 0, 0), (500, 500, 50))

def load_annotation(annotation_file: Path) -> None:
    # Read the WEBKNOSSOS annotation file (a zipped WKW)
    annotation = wk.Annotation.load(annotation_file)

    # Treat it as a regular WK volume layer
    with annotation.temporary_volume_layer_copy() as segmentation_layer:
        # Do any standard layer operation, e.g. reading a cutout as a numpy array
        mag_view = segmentation_layer.get_finest_mag()
        segments =

        # Write segmentation IDs to an OME Tiff file
            segments.T,  # note, the tiff lib use different channel order
                "axes": "ZYXC",

if __name__ == "__main__":
    # Path to annotation file on your computer
    ANNOTATION_FILE = Path("")