siemdejong / hhg-preprocessing

Preprocessing pipeline for higher harmonic generation images for use with AI

hhg-preprocessing

Preprocessing pipeline specifically for higher harmonic generation images for use with AI.

Currently, this repo implements

• Tissue segmentation

Install

git clone https://github.com/siemdejong/hhg-preprocessing
cd hhg-preprocessing
pip install -e .

Tissue segmentation

To use the reimplemented EntropyMasker tissue segmentation algorithm

import numpy as np
from PIL import Image
from hhg_preprocess import entropy_masker
image = np.array(Image.open("path/to/your/image.png"))
mask = entropy_masker(image)

or use the CLI:

entropy_masker --data <data/data.txt> --output_dir <output_dir>

where <data/data.txt> is a text file with input filenames separated by newlines. If desired, one can change the default settings in hhg_preprocess.entropy_masker.config.default.py.

Data Version Control

To start a DVC pipeline, populate <data/data.txt> with filenames and execute

dvc repro

For more on DVC, refer to the DVC documentation

hhg-preprocessing's Issues

Add tissue packing

Tissue can be packed if more fragments are in one image. Saves space.