Improc

A collection of image processing routines.

Installation

git clone https://github.com/fmi-basel/improc.git
pip install improc/

Usage

This package consist of a collection of somewhat common functions that are not directly available in libraries such in scikit-image or opencv. That includes (non-exhaustive):

Parsing file collections

The io sub-module provides a tool to parse a collection of files following the python string formatting syntax and saves the result in a pandas DataFrame. An accessor is also provided to expand pandas with read/write shortcuts for images and csv files. see parse_collection.ipynb for more details.

Cleaning up labels

label and morphology sub-modules contain routines to clean up segmentation labels (filter size, fill-holes, edge smoothing, etc.)

Converting 3D anistotropic labels to smooth vtk meshes

see mesh sub-module

Finding objects in 3D stacks from z, x and y projections

see segmentation sub-module

Extracting features

The regionprops sub-module extends scikit-images' regionprop to multi-channel, multi-label data in a modular way.

Resampling intensity images and labels to match a given voxel size

see resample sub-module

Reading multi-scene czi files

The czi_reader sub-module provides draft utility functions to read multi-scene czi files. If working prefere using czifile or more actively developed aicspylibczi. (both attempt to create a giant image containing all the scenes and run out of memory)