Contributed Nodes

These are modules that are not central to the functioning of MorphoCut, but are nevertheless useful in its application context.

EcoTaxa

Read and write EcoTaxa archives.

“EcoTaxa is a web application dedicated to the visual exploration and the taxonomic annotation of images that illustrate the beauty of planktonic biodiversity.”

class morphocut.contrib.ecotaxa.Archive(archive_fn, mode='r')

A generic archive reader and writer for ZIP and TAR archives.

read_member(member_fn)

Raises:

MemberNotFoundError – if a member was not found

class morphocut.contrib.ecotaxa.EcotaxaObject(meta, image_data, archive_fn=None, index_fn=None, default_mode=None)

meta

Object metadata

Type:

Mapping

index_fn

Source index filename inside the archive.

Type:

str, optional

archive_fn

Sourche archive filename.

Type:

str, optional

default_mode

Default mode for image loading. See PIL Modes.

Type:

str, optional

get_image(img_rank=None, mode=None)

Get an image for the object.

Parameters:

• img_rank – Rank of the image. (Default: The first one.)

• mode – Mode to load the image. See PIL Modes.

morphocut.contrib.ecotaxa.EcotaxaReader(archive_fn: str | Variable[str], *, query: str | None | Variable[str | None] = None, prepare_data: Callable[[pd.DataFrame], pd.DataFrame] | None = None, verbose=False, keep_going=False, print_summary=False, encoding='utf-8', index_pattern='*ecotaxa_*', columns: List | None = None, image_default_mode=None)

Stream Read an archive of images and metadata that is importable to EcoTaxa.

Parameters:

• archive_fn (str, Variable) – Location of the archive file.

• query (str, Variable, optional) – A query string to select a subset of the data. See pandas.DataFrame.query().

• prepare_data (callable, optional) – A function that receives a pandas.DataFrame and returns a modified version.

Returns:

object – An EcotaxaObject that allows to access image(s) and metadata.

To read multiple image ranks, provide a tuple of ints as img_rank. The first output will then be a tuple of images.

The TSV file needs at least an img_file_name column that provides the name of the image file. Other columns are read from meta.

The TSV file MAY contain a row of types after the header ("[f]" for numeric columns, "[t]" else).

Example

with Pipeline() as p:
    obj = EcotaxaReader("path/to/archive.zip")
p.transform_stream()

morphocut.contrib.ecotaxa.EcotaxaWriter(archive_fn: str | Variable[str], fnames_images: Tuple[str, ...] | Variable[Tuple[str, ...]] | List[Tuple[str, ...] | Variable[Tuple[str, ...]]], meta: Mapping | Variable[Mapping] | None = None, object_meta: Mapping | Variable[Mapping] | None = None, acq_meta: Mapping | Variable[Mapping] | None = None, process_meta: Mapping | Variable[Mapping] | None = None, sample_meta: Mapping | Variable[Mapping] | None = None, meta_fn: str | Variable[str] = 'ecotaxa_export.tsv', store_types: bool = True)

Create an archive of images and metadata that is importable to EcoTaxa.

Parameters:

• archive_fn (str) – Location of the output file.

• fnames_images (Tuple, Variable, or a list thereof) – Tuple of (filename, image) or a list of such tuples. filename is the name in the archive. image is a NumPy array. The file extension has to be one of ".jpg", ".png" or ".gif" to meet the specifications of EcoTaxa.

• meta (Mapping or Variable, optional) – Metadata to store in the TSV file. Each key corresponds to a column in the resulting file.

• object_meta (Mapping or Variable, optional) – Metadata stored with object_ prefix.

• acq_meta (Mapping or Variable, optional) – Metadata stored with acq_ prefix.

• process_meta (Mapping or Variable, optional) – Metadata stored with process_ prefix.

• sample_meta (Mapping or Variable, optional) – Metadata stored with sample_ prefix.

• meta_fn (str, optional) – TSV file. Must start with ecotaxa.

• store_types (bool, optional) – Whether to add a row with types after the header. Defaults to True, according to EcoTaxa’s specifications.

If multiple images are provided, image and image_name must be tuples of the same length.

The TSV file will have the following columns by default:

• img_file_name: Name of the image file (including extension)

• img_rank: Rank of image to be displayed. Starts at 1.

Other columns are read from meta. The file will contain a column for each object in the stream.

Example

with Pipeline() as pipeline:
    image_fn = ...
    image = ImageReader(image_fn)
    meta = ... # Calculate some meta-data
    EcotaxaWriter("path/to/archive.zip", (image_fn, image), meta)
pipeline.transform_stream()

See also

For more information about the metadata fields, see the project import page of EcoTaxa.

exception morphocut.contrib.ecotaxa.MemberNotFoundError

class morphocut.contrib.ecotaxa.TarArchive(archive_fn, mode='r')

read_member(member_fn)

Raises:

MemberNotFoundError – if a member was not found

exception morphocut.contrib.ecotaxa.UnknownArchiveError

class morphocut.contrib.ecotaxa.ZipArchive(archive_fn, mode='r')

read_member(member)

Raises:

MemberNotFoundError – if a member was not found

ZooProcess

Feature calculation like in ZooProcess.

Zooprocess is a suite of routines in ImageJ macro language for Plankton image analysis.

morphocut.contrib.zooprocess.CalculateZooProcessFeatures(regionprops: T | Variable[T], meta: dict | Variable[dict] | None = None, prefix: str | Variable[str] | None = None)

Calculate descriptive features similar to ZooProcess using skimage.measure.regionprops().

Parameters:

• regionprops (RegionProperties or Variable) – RegionProperties instance returned by FindRegions.

• meta (dict or Variable, optional) – Meta-data dictionary to update.

• prefix (str or Variable, optional) – Prefix for all keys.

Example

with Pipeline() as p:
    image = ...
    mask = ...

    regionprops = FindRegions(mask, image)

    features = CalculateZooProcessFeatures(regionprops)

morphocut.contrib.zooprocess.regionprop2zooprocess(prop)

Calculate zooprocess features from skimage regionprops.

Notes

• date/time specify the time of the sampling, not of the processing.