ICA on Mixed Images

Usage of ICA to separate two mixed images, based on Nathan Kutz lecture.

About this project

Implementation of ICA (a.k.a. Independent Component Analysis) to separate two images which were mixed. This code is based on the lecture of Nathan Kutz, professor of Applied Math at Washington University. For this technique to work, it is imperative that all original components to be INDEPENDENT AND NON GAUSSIAN, so that they can be separated.

Getting Started

Prerequisites

To run the algorithms in this repo, you'll need to have Python 3 installed.

Python dependencies

To run the notebook, you'll need to import all of the libraries below:

$ pip3 install numpy

$ pip3 install matplotlib

$ pip3 install opencv-python

$ pip3 install jupyter-notebook

References

1. Independent Component Analysis 1. J Nathan Kutz, professor of applied math. Washington University. Available at: https://www.youtube.com/watch?v=_e4SN4TWlgY. Access in April, 2019.

2. Independent Component Analysis 2. J Nathan Kutz, professor of applied math. Washington University. Available at: https://www.youtube.com/watch?v=olKgmOuAvrc. Access in April, 2019.

3. Independent Component Analysis 3. J Nathan Kutz, professor of applied math. Washington University. Available at: https://www.youtube.com/watch?v=Ad6kMhJbqoY. Access in April, 2019.