NIC@HMS python2: Intro to image processing

https://docs.conda.io/en/latest/miniconda.html

If you haven't already, add conda forge to your channels

conda config --add channels conda-forge
conda config --set channel_priority strict

tip: to download a version of miniconda with conda-forge already setup, use miniforge

# to clone
git clone https://github.com/tlambert03/hms_pyintro2.git

or download and unzip: https://github.com/tlambert03/hms_pyintro2/archive/refs/heads/master.zip

... then cd into the new directory

cd hms_pyintro2

conda create -n imgproc python
conda activate imgproc

note: anytime you close and reopen your terminal, you'll need to reactivate your conda environment with conda activate <name_of_env>

pip install -r requirements.txt

this will pip install each of the requirements listed in the requirements.txt file. You could also install them with conda if you'd like. See the previous lecture for details on pip vs conda.

jupyter-lab 01_image_io.ipynb

• pathlib.Path for filesystem operations. (offers similar functionality to the os.path module, in an object-oriented design)
• Open files (like text files) with open, or Path.read_text
• Loading image data:
  • tifffile - read and write TIFF files. Used internally by many other packages.
  • imageio - read and write a wide range of image data (but use volread instead of imread for nd tiffs)
  • scikit-image io module - Utilities to read and write images in various formats. Wraps tifffile and imageio, and other plugins.
  • pims - consistent interface for image sequences. offers Bioformats wrapper, and pure-python nd2 readers with nd2reader and pims_nd2.
  • so many others... let me know if you have a challenging image format...
• Viewing images:
  • matplotlib.pyplot.imshow excellent image viewer with lots of options... but mostly designed for 2D data.
  • napari.view_image view n-dimensional arrays in the napari viewer.
  • napari.view_path, open file path in napari viewer. Can install use napari plugins to read proprietary formats.
• Numpy - the fundamental package for scientific computing in Python. It's critical to understand the basics of working with numpy arrays when dealing with image data (indexing, slicing, operations, etc...). Start with the quickstart and go from there...
• Saving images:
  • tifffile.imsave
  • skimage.io.imsave
  • numpy.save

• Start with scikit-image for most of your basic image processing and analysis needs.
  • Start with the user-guide
  • tutorial on scikit-image with napari
• scipy.ndimage has lots of good functions (filtering, interpolation, measurements) for nd images.
• OpenCV has a ton of functionality.
• napari tutorials
  • napari training course presented for NEUBIAS webinar.