bm_dataset.rescale_tissue_images module

Converting images to particular scales

With given path pattern to images crete particular scales within the same set

Note

Using these scripts for 1+GB images take several tens of GB RAM

Sample usage:

python rescale_tissue_images.py         -i "/datagrid/Medical/dataset_ANHIR/images_private/COAD_*/scale-100pc/*.png"         --scales 5 10 25 50 -ext .jpg --nb_workers 4

Copyright (C) 2016-2019 Jiri Borovec <jiri.borovec@fel.cvut.cz>

bm_dataset.rescale_tissue_images.arg_parse_params()[source]

parse the input parameters :return dict: parameters

bm_dataset.rescale_tissue_images.main(path_images, scales, image_extension, overwrite, nb_workers)[source]

main entry point

Parameters

  • path_images (str) – path to input images

  • scales (list(float)) – define scales in percentage, range (0, 100)

  • image_extension (str) – image extension used on output

  • overwrite (bool) – whether overwrite existing image on output

  • nb_workers (int) – nb jobs running in parallel

Returns

bm_dataset.rescale_tissue_images.scale_image(img_path, scale, image_ext='.jpg', overwrite=False)[source]

scaling images by given scale factor

Parameters

  • img_path – input image path

  • scale (int) – selected scaling in percents

  • image_ext (str) – image extension used on output

  • overwrite (bool) – whether overwrite existing image on output

bm_dataset.rescale_tissue_images.wrap_scale_image(img_path_scale, image_ext='.jpg', overwrite=False)[source]