A meta plugin for processing timelapse data in napari timepoint by timepoint
A meta plugin for processing timelapse data timepoint by timepoint. It
enables a list of napari plugins to process 2D+t or 3D+t data step by step when the user goes
through the timelapse. Currently, these plugins are using
napari-time-slicer enables inter-plugin communication, e.g. allowing to combine the plugins listed above in
one image processing workflow for segmenting a timelapse dataset:
The workflow can then also be exported as a script. The 'Generate Code' button can be found in the Workflow Inspector
If you want to convert a 3D dataset into a 2D + time dataset, use the
Tools > Utilities > Convert 3D stack to 2D timelapse (time-slicer). It will turn the 3D dataset to a 4D datset
where the Z-dimension (index 1) has only 1 element, which will in napari be displayed with a time-slider. Note: It is
recommended to remove the original 3D dataset after this conversion.
Working with large on-the-fly processed datasets¶
Using the napari-assistant complex image processing workflows on timelapse datasets can be setup.
In combination with the time-slicer it is possible to process time-lapse data that is larger than available computer memory.
In case the workflow only consists of images and label-images and out-of-memory issues arise, consider storing intermediate results on disk following this procedure:
After setting up the workflow and testing it on a couple of selected frames, store the entire processed timelapse dataset to disk
using the menu
Tools > Utilities > Convert to file-backed timelapse data (time-slicer). It will open this dialog, where you can select
It is recommended to enter a folder location in the text field.
If not provided, a temporary folder will be created, typically in the User's temp folder in the home directory.
The user is responsible for emptying this folder from time to time.
The data stored in this folder can also be loaded into napari using its
File > Open Folder... menu.
Executing this operation can take time as every timepoint of the timelapse is computed. Afterwards, there will be another layer available in napari, which is typically faster to navigate through. Consider removing the layer(s) that were only needed to determine the new file-backed layer.
Usage for plugin developers¶
Plugins which implement the
napari_experimental_provide_function hook can make use of the
@time_slicer. At the moment,
only functions which take
napari.types.LabelsData and basic python types such as
float are supported. If you annotate such a function with
@time_slicer it will internally convert any 4D dataset
to a 3D dataset according to the timepoint currently selected in napari. Furthermore, when the napari user changes the
current timepoint or the input data of the function changes, a re-computation is invoked. Thus, it is recommended to
only use the
time_slicer for functions which can provide [almost] real-time performance. Another constraint is that
these annotated functions have to have a
viewer parameter. This is necessary to read the current timepoint from the
viewer when invoking the re-computions.
import napari from napari_time_slicer import time_slicer @time_slicer def threshold_otsu(image:napari.types.ImageData, viewer: napari.Viewer = None) -> napari.types.LabelsData: # ...
You can see a full implementations of this concept in the napari plugins listed above.
If you want to combine slicing in time and processing z-stack images slice-by-slice, you can use the
Make sure, to insert it after
@time_slicer as shown below and implemented in napari-pillow-image-processing
from napari_time_slicer import slice_by_slice @time_slicer @slice_by_slice def blur_2d(image:napari.types.ImageData, sigma:float = 1, viewer: napari.Viewer = None) -> napari.types.LabelsData: # ...
You can install
napari-time-slicer via pip:
pip install napari-time-slicer
To install latest development version :
pip install git+https://github.com/haesleinhuepf/napari-time-slicer.git
Contributions are very welcome. Tests can be run with tox, please ensure the coverage at least stays the same before you submit a pull request.
Distributed under the terms of the BSD-3 license, "napari-time-slicer" is free and open source software
If you encounter any problems, please file an issue along with a detailed description.
- 10 July 2022
- 12 November 2021
- Information not submitted
- Stars: 1
- Forks: 3
- Issues + PRs: 3