A wrapper for Ben's LBP bio pipeline
Project description
ImageTextureFinder
A project to create an easy-to-use way of finding areas of common patterns and structures within an image. Should work on any image, designed for use on any biological images including DAPI, IMC and H&E.
See sample_run.sh
for details.
- Branch
baseline
is the most stable. It is ready for pip packaging. - Branch
pip
is stale at the moment - Branch
dev
is unstable and for dev purposes only.
Container
Image tag is mkrooted/imbg-fastlbp
. Hosted on Docker Hub (https://hub.docker.com/repository/docker/mkrooted/imbg-fastlbp/general).
See https://github.com/imbg-ua/fastLBP-sandbox for details
Guides
How to build and deploy a pip package
Src: https://packaging.python.org/en/latest/tutorials/packaging-projects/
- Add your access token to
.pypirc
# ~/.pypirc [pypi] username = __token__ password = pypi-TOKEN_FROM_YOUR_PYPI_SETTINGS_GOES_HERE
- Ensure that your Python is 3.8 because the package targets Python 3.8 and thus requires to be build using this Python version
python --version # Should show Python 3.8.something
- Install prerequisites (
twine
andbuild
)pip install --upgrade twine build
- Edit project version in
pyproject.toml
- Build and upload the project
# while in root project directory python -m build # .whl and .gz output will be at ./dist directory python3 -m twine upload dist/* # note that this can accidentally upload unneeded builds
Algorithm notes
Step 1 performs an LBP and creates histograms for each method.
Method is a combination of the following parameters:
- image name
- image channel
- LBP radius
- LBP number of points
Every method's result got saved into the separate .npy
file. There is a correspondence betweeen a method and a computational job.
Step 2 collects all the results and concatenate them along the features dimension. That means that feature vector of a patch is a concatenation of all LBP codes from all channels and all LBP radii.
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for fastlbp-baseline-imbg-0.0.7.dev1.tar.gz
Algorithm | Hash digest | |
---|---|---|
SHA256 | 83fab3bbe808c7dbd25c99a579b9c697c81ec8353cb6124cc51e282ad1e6f490 |
|
MD5 | 78a31a37cc7406908e59c4a23337414b |
|
BLAKE2b-256 | 5cfba4d4aadd41b5d3361cf33b97f41887b06c2b3563e8687288a21131c74295 |
Hashes for fastlbp_baseline_imbg-0.0.7.dev1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 2d81c94ce47e0181aecf272f2cb548314ad3d491740227243118eb23ce7e56b7 |
|
MD5 | f3d21d84205d0b82ed8c4c35c0c59ea3 |
|
BLAKE2b-256 | a20e1baae48790933a08f5760df6ea5d40868588f26813c73a7b2763db7c992e |