A Machine Learning-based Optical Fine Alignment tool, created as a R&D project under the Vacation Students' Programme (VSP-2019) of Inter-University Centre for Astronomy and Astrophysics (IUCAA).
This project was performed under the supervision of Dr. Sreejith Padinhatteeri (https://github.com/sreejithpa)
-
In a complex optical system, the identification of source of misalignment is always a difficult task, and is generally performed by the experienced optical engineers/scientists.
-
The project aims to model the variations in the Zernike Coefficients due to the misalignment of the optical components of the telescope, and then apply various Machine Learning algorithms to accurately predict the exact component having the misalignment and the actual misalignment.
A presention with all the implementation details is provided here
This project is currently licensed under The MIT License (MIT). A copy of LICENSE should be present along with the source code.