version 1.1.1

• fixed wsidataset for pytorch (57c2376)
• hotfix of an argument of wp.annotation (66201c4)

version 1.1.0

• on_annotation parameter can be set flexibly (fa915e5, b583731)
• save images as jpg instead of png (fd22b9b)
• verbosity state fixed (d605591)
• NDP.View2-styled annotation enabled (2afa343, 0c2f6d8)

version 1.0

• mask can be resized of RAM is not enough
• wsiprocess.rule.Rule can take also as a dict
• the magnification setting is enabled
• result coordinations are saved as csv
• geojson styled annotation is enabled
• pytorch dataset helper implemented
• default output extension changed to jpg
• documentation page overhauled

version 0.9

• foreground extract function (annotation.make_foreground_mask) can take lambda functions

version 0.8

• Method name "none" is switched with "evaluation"
• verbose option is available for showing patching progress

  wsiprocess evaluation sample.tiff --verbose # or -ve
  

• Backend is switched with openslide. If pyvips is set to be the backend in slide.py, wsiprocess handles WSIs with pyvips.

version 0.7

• only if --extract_foreground is set, wsiprocess extracts patches from the WSI
• translate dot annotation to a small bounding box
• wsiprocess can handle SlideRunner annotations (bnapora)
• wsiprocess can handle QuPath annotations

version 0.6

• Work with custom function for foreground estimation.
• Foreground class can be extracted from command line.

version 0.5

• FIxed find_bbs to find all the bounding boxes meeting the conditions.
• Dockerfile updated to run wsiprocess in a docker container.
• annotation.Annotation.make_foreground_mask can take custom thresholding functions.
• Examples in the README.md significantly changed.
• Installation for Linux updated.

version 0.4

• arguments of cli.py updated
  • the order of "method" and "wsi" is switched
  • mm is added; Annotation can make mask from the minimum and the maximum value.
  • no_patches enabled, and extract_patches disabled; without setting any args about extracting patches, wsiprocess extracts patches.
• Annotation.from_image()
  • Annotation can make mask from a image.
• Citation updated

version 0.3

Update

• added wsiprocess.annotation.Annotation.from_image to load mask data from image.
• release added for DOI@zenodo