This project was done in cooperation with Professor Firas Khasawneh .
The ultimate goal for this project is to use machine learning classifiers on objects detected in space based on their light reflection. The big challenge to train such a classifier is to gather enough training data.

As it is unfeasable for me to gather real-world reflection data for thousands of space objects, I spend the majority of last semester building a simulation software that will randomly calculate possible orbits of space objects. For each orbit the light reflection is simulated at each time step based on the shape of either debris or a satellite.
This data was then used to train a Discrete Frechet Distance (DFD) algorithm. In the fall semester, I will be working on using different machine learning algorithms in order to improve our classifier.

For more details regarding our research project please refer to the following pdf: