KerrPy

An openCV based Python project for processing Kerr microscopy image. KerrPy is used to extract ellipse fits of bubble domains created under the application of out-of-plane and in-plane fields. These fits are used to estimate the underlying Dzyaloshinskii Moriya interaction (DMI)

Installation requirements

• Anaconda (preferably 64 bit) for managing Python environments.

• A Python environment

• Installation of dependencies

Python environment : envKerr

Every Python project is recommended to have a separate python environment. There are two ways to create a new environment.

Through Navigator

A walkthrough of the installation is given in YouTube video of installation walkthrough

• Activate the environment

• Install packages in the environment

Packages

core packages	image processing packages	IDE packages
numpy	opencv	spyder
scipy
matplotlib

Anaconda maintains "channels" for storing repositories. Core packages are available in standard/default channel. So can be directly downloaded whereas opencv is only present in "conda-forge" channel.

Update: The default channel also provides opencv.

conda-forge is a community maintained channel. The community is comprised mostly of researchers. So we need to add conda-forge channel to the environment.

Newer versions of Anaconda has opencv in the default channel So no need to add conda-forge However latest versions of opencv may not be supported in the default.

However we will install the default channel version.

command line way

• through command line. Use the environment.yml in the top level directory.

    conda env create -f environmentKerrPy.yml
  

• activate the environments

    conda activate envKerr
  

Configuration

• As we need an interactive backend to select a Region of Interest (ROI), Spyder has to be configured to use the Qt5 backend.

  Tools > Preferences > IPython Console > Graphics > Qt5 backend

• Then the first file to configure is the globalVariables.py. It has the following categories of parameters

  • Training parameters,
  • File handling parameters,
  • Flag parameters,
  • Matplotlib parameters, and
  • Image parameters

These parameters are self-explanatory. For additional information, refer to the adjoining comments.

Execution

Now run the scripts in the following order

• processDataWithCustomROI.py
• processFigures.py
• processVelocity.py

Output

The output is piped to the PostProcessing folder. The ellipse fit(yellow) is shown below along with the inliers (green) and outliers (red).

The restored image is shown below.