-
Notifications
You must be signed in to change notification settings - Fork 0
Open
Labels
enhancementNew feature or requestNew feature or request
Description
We should switch our image loading to bioio. We currently use imageio which is fine for simple tiff files but will not work for anything beyond that.
Bioio has the following advantages:
- Can read shape and dtype without loading any image data (to make our core MetaData).
- In the future we can use its lazy loading using dask arrays (like zarr)
- Is designed to load proprietary image formats. We will start with
tiffbut eventually need drag-drop support of commercial scope file formats like Olympus (oir) , Zeiss (czi), Nikon (nd2), etc, etc - Is maintained on GitHub and we can help develop the code and get questions answered. I followed these folks from their original package Allen Institute for Cell Science (aicsimageio) and trust their expertise.
It uses a plugin system to extend itself to new file formats. For example to load a tiff, you need the plugin
pip install bioio
pip install bioio-tifffileThen in Python
from bioio import BioImage
path = '/Users/cudmore/Sites/PyMapManager-Data/maps/rr30a/rr30a_s0_ch1.tif'
img = BioImage(path)
img.shape # (1, 1, 70, 1024, 1024)
img.dtype # dtype('uint16')
# using Dimensions
img.dims # <Dimensions [T: 1, C: 1, Z: 70, Y: 1024, X: 1024]>Finally, we can write our own bioio plugin to load our mmap zarr format!!! Hold off on this for now!!!
Metadata
Metadata
Assignees
Labels
enhancementNew feature or requestNew feature or request